The BreadLab Flour Reference

This reference addendum is the technical foundation for the book: when a chapter mentions ash, or extraction, or why a flour absorbs more water than its label-stated protein number predicts, this is where the explanation goes so that the baking chapters can stay focused on baking (and my opinions about how best to approach it). Though fascinating, you do not need to read this information in its entirety unless you are as nerdy as I, but it's a glimpse into a source of my opinions and thought processes, and it's here to serve as a reference when you want to know not just what a flour is, but what technical and other information informed the way I approach it.

A Note About BreadLab Observations

Throughout this reference you'll see material identified as a BreadLab Observation. These are not based upon purely scientific studies, and they shouldn't be interpreted as scientific proof. If pressed I'd refer to them as semi-scientific anecdotal information. I'm not conducting double-blind experiments, publishing peer-reviewed research, or working with sample sizes large enough to establish statistically significant conclusions.

That said (and admittedly), I do try to use a very consistent baking process and change as few variables as possible so I can observe how different flours, techniques, and decisions affect the bread, so that over time, I can see what patterns emerge. Those patterns inform my thinking and my recommendations, but they remain personal observations rather than empirical proof. Of course, things like handling and shaping are hardest to repeat consistently, let alone quantify, but I have made bread so many times that I consider muscle memory a decent proxy for scientific rigor.

Moreover, a central thesis of my method is that careful observation has always been part of good baking. Every experienced baker develops opinions by paying attention to what happens repeatedly on the bench and in the oven. BreadLab is simply my attempt to make those observations more systematic, more transparent, and more useful to other bakers, and my app is my attempt to provide the exact tool you will need to do just that: observe, compare, and learn from your experiences.

So, when I describe a BreadLab Observation, think of it as evidence that has both informed and been informed by my personal judgment. You are by all means encouraged to test those observations in your own kitchen, challenge them, and draw your own conclusions. If I am wrong, I want to know and learn, same as you. I just ask that you take on faith that I'm trying to bridge the gap between cereal science and everyday baking in a way that is most helpful to bakers of every experience level.

The Provenance Model

One thing I hope sets this reference apart from other writing about flour: my claims usually tell you where I came by that opinion or observation. Online advice is usually a toxic blend of hard chemistry, industry norms, personal preferences, semi-true old saws and old wives' tales, and outright myths, combined into a single confident voice, and the reader is often left with no way to tell which is which.

I will make every attempt to always let you know the source of my work and when something is just based on experience or even a hunch, though he point is not to hedge. I am never shy about telling you what I think you should do, but it is only fair that you know whether you're reading settled chemistry, standard industry practice, something I've watched happen on my own bench with my best explanation for why it happens, or simply my opinion (which I always explain most consistently) and you can weigh each accordingly. I think that transparency makes the advice more trustworthy, not less, because I'm never asking you to accept something just because I said it, and I will provide enough information that you can repeat the test and see for yourself just how true or universal my conclusions actually are.

Part I — Understanding Wheat

Everything a flour does in your hands and in the bowl begins with the qualities of the wheat, but it doesn't end there. Besides the variety of wheat, how it's milled and how it's screened, sifted, fortified and otherwise handled afterward all shape the flour that ends up in your dough.

Wheat Classes

If you've baked with more than one type of flour, you've felt how different two "bread flours" can be, and a lot of that difference is established by what class it is before the wheat ever reaches the mill. Wheat generally gets sorted by three traits: when it was planted (spring or winter), how hard the kernel is (hard or soft), and the color of its bran (red or white). Hardness matters most to a bread baker because hard wheats have more and stronger gluten-forming protein.

Those three traits determine membership in a handful of commercial classes, and knowing which class a flour came from often tells you more than the protein number on the bag.

The pattern: hard wheats build structure, soft wheats build tenderness, and red versus white is mostly flavor (red bran carries a faint bitterness that white bran lacks). Durum is an outlier, hard as flint but with highly extensible gluten that stretches endlessly without springing back, making it ideal for pasta.

Kernel Anatomy

Extraction rate tells you how much of a wheat kernel ended up in your flour. But to understand the practical effect of that number, you need to understand the three parts of the kernel, because each one behaves differently in dough. A wheat kernel, or wheatberry, is botanically a fruit called a caryopsis (a single seed whose coat is fused to the fruit wall) and it has three edible parts protected by an inedible husk that's knocked loose by the combine's thresher, winnowing and then discarding the chaff.

Endosperm: The bulk of the kernel and the source of white flour. It's mostly starch, plus the proteins that form gluten when they meet water. Refined white flour is essentially pure endosperm, though it's whitest when bleached.

Bran: The hard outer layers, packed with fiber and minerals. In dough, bran is a mixed blessing: it brings flavor, color, and nutrition, but its sharp-edged particles interrupt the gluten network, compete for water, and speed fermentation. More bran means more taste and with it, more complication.

Germ: The embryo of the fruit, small, but rich in oils, enzymes, and vitamins, which is why it carries much of the flavor and nutrition, but also most of the reason whole-grain flour goes rancid: the unsaturated oils oxidize quickly once the kernel is broken and the germ meets the air, which is why wheat is best stored in berry form until milling is required.

The rough proportions — around 83% endosperm, 14% bran, 3% germ — are worth knowing, though they vary by (and to a large extent define) cultivars. For baking, the approach is simpler: white flour is the endosperm alone, whole grain is all three, and knowing the makeup of your flour allows you to control flavor, texture, aroma, and other qualities in your bread by knowing what happens in the extraction rates between white, starting at 72%, and whole grain, at 100%.

Protein Percentage

Sooner or later, every serious baker compares commercially available flours, both of which say 12.5% nominal protein, but one makes a supple, strong dough that shapes like a dream and the other feels sticky, slack, or fights stretching like a cold rubber band. It's the same number, but very different resulting bread. The number on the bag is protein quantity, and quantity is important, but only half the story.

Protein percentage tells you roughly how much structural raw material a flour holds, a real and useful clue because higher-protein flours can generally build more gluten than lower-protein ones. But in my standardized BreadLab evaluations, dough performance on the bench has not tracked directly to protein percentage nearly as closely as you'd expect. In my testing of scores of flours and grains, two flours at exactly the published number routinely handled nothing alike, and in fact a lower-protein flour has more than once outperformed a higher-protein one on the bench and in the final crust and crumb report.

Protein Quality

When two flours share a protein number but behave completely differently, some factor other than protein quantity is obviously at work. That something is protein quality: the behavior and balance of the two gluten-forming proteins, glutenin and gliadin. When flour meets water, these combine to form gluten. Glutenin provides elasticity and tenacity — the dough's ability to spring back and hold its shape. Gliadin provides extensibility — the ability to stretch without tearing. The ratio between them, set largely by wheat genetics, decides whether a dough is strong and bouncy, slack and stretchy, or balanced in a way you actually want. This affects my enjoyment of working with a given flour, even when the results are not nearly as straightforward in the resulting bread.

This is also why heirloom or ancient wheat can boast a high protein number and still make a weak, fragile dough. Though plentiful, the protein may simply not include the strong polymeric glutenin that modern bread wheats were bred to carry.

Gluten and the Role of Bran

While gluten is created through hydration, it's strength in the dough has to be developed and built into a useful structure. The proteins sit inert in the flour until water and agitation bring them together into an elastic, gas-trapping web that lets bread rise instead of spreading, and while the resulting elasticity holds the shape, the extensibility lets it expand. Good bread is made by achieving balance, and that balance is what mixing, folding, and time are what you are trying to manage on your bench.

Bran complicates the matrix that gluten needs to form for strength, but the common shorthand "bran's sharp edges cut gluten" is too simplistic to be useful. While bran particles do interrupt the network, they also compete for water, change how the dough feels and flows, reduce extensibility, and add extra amylase enzyme activity. Which effect dominates the effect depends on the size of the bran particles, how much bran is present, the hydration, the fermentation time, and the dough itself. Fine bran isn't automatically harmless, and coarse bran isn't automatically ruinous, but as part of the system they must be taken into account.

Ash Content

Ash content in much like caloric content: you can only know that actual number by the indirect method of burning the thing you are trying to measure and looking at an output of the process. It's the same with extraction rate. If you want to know how much of the outer kernel is left in a bolted flour, there's an indirect measurement that gets you closer than just weighing the flour before and after bolting, and it's called ash. It matters because everything the outer kernel brings to bread — flavor, color, minerals, fermentation speed, thirst — is based on how much of that outer kernel is present, and ash is the cleanest single clue to that amount. Burn a flour sample in a laboratory furnace until every scrap of organic material is gone, and the ash residue that remains is pure mineral, and because minerals predominately are found in the bran and germ rather than the endosperm, more ash means more of the outer kernel survived into the flour and ended up as ash. As a result, very white flour runs low, ash numbers of around 0.4–0.5%; high-extraction flour runs higher, and whole grain runs highest of all.

One of the most common false equivalencies in flour specification worth knowing is that ash is not the same measurement as extraction, because though two numbers are the same, the component parts or the ash can be different have different effects on the flour. Two mills can reach the same ash number and end up with different extraction percentage, protein levels, flavor, and behavior, and two flours can share the exact same extraction rate and differ in ash residue percentage. Ash is a good clue, an indirect measurement of how much of the kernel is in the bag, but it is not a direct one.

Extraction Rate

Of all the numbers that describe a flour, extraction rate predicts the most about how it will behave — but though common practice in commercial products and some specialty artisan flours, most American consumer products actually don't provide it at all. Extraction is there to tell you how much of the original cleaned whole grain berry ended up as flour?

Extraction Rate Percentage = (Weight of Flour / Weight of Grain)

100 kilograms of grain yielding 75 kilograms of flour is 75% extraction. A low extraction number (65–75%) means the mill kept only the clean inner endosperm and discarded the outer layers. A high number (80%+) means more of the bran and germ stayed in. As extraction climbs, so do mineral content, ash, color, flavor, water demand, and fermentation speed, as do complications, because more bran interferes with gluten and the germ's oils shorten shelf life.

A few extraction landmarks come up again and again, so they're worth seeing.

Patent flour at ~72% is the reliable white bread flour most bakers start with, trading flavor for predictability, easy handling and mild, sweet flavor. The 80–85% band is where a lot of the most interesting bread is found, keeping enough strong, pure white endosperm to build real structure while carrying the minerals, flavor, and fermentation energy of the outer grain. Whole grain gives you everything the wheat has, and every complication that comes with it. None is better than the others; they're just different tools requiring different methods. One caution: extraction is not protein, not ash, and not particle size. They often move together, but any one can vary while the others stay the same. A 13.5% protein flour at 72% extraction and a 13.5% protein flour at 100% extraction share a number on the bag and behave like two entirely different ingredients in your dough.

Falling Number

Flour is alive with alpha- and beta-amylase enzymes whose job in the symbiotic culture in a starter is to break complex starches into the simple sugars that feed yeast's fermentation. Too little enzymatic activity and dough ferments sluggishly, browns poorly, and tastes flat. Too much, and it turns sticky, gummy, and weak. Bakers rarely measure enzymes directly, but mills do, and the measurement is sometimes posted on a commercial or specialty artisanal product's bag or website, so it's worth knowing what it means.

To establish a Falling Number, a weighted plunger is timed as it drops through a warmed slurry of flour and water. More enzyme activity thins the slurry, so the plunger falls faster and the number is lower. The preferred zone is roughly 250–300 seconds, but above about 400, activity is low, which can result in slow fermentation, pale crust, and little caramelization due to the lower available sugar content active enzymes would free. Below about 200, often a sign of grain that sprouted in the field before harvest, activity is excessive, which can result in sticky dough, a gummy crumb, poor spring, and bread that stales fast because the enzymes are chewing through the starch that holds it up.

Because modern harvests are selected for sound, unsprouted grain, commercial flour often runs naturally low in amylase, and mills correct this by adding a tiny fraction of diastatic malted barley flour or fungal amylase to reach the target zone. Supplementing with malt due to a poor falling number is covered in Part II.

Part II — Understanding Flour

The same wheat can become several very different flours depending on what type of processing the mill uses, both in the milling and the bolting processes, and this is where a lot of a flour's real personality is decided. Two bags from the same field can behave differently if one was roller-milled and refined while the other was stone-ground whole.

Roller Milling

Almost all commercial flour is made using roller mills, and understanding the process explains why two flours at the same extraction can still differ so much. The mill doesn't only make one flour, it makes many and even choses which to combine to create a specific product. Roller milling passes grain through a sequence of steel cylinders that shear the kernel apart in stages, breaking, scraping, grinding, and sifting. What comes out isn't a single of flour particles, but dozens, each differing in particle size, mineral content, bran, protein concentration, and starch damage. The miller then recombines selected streams into a finished flour at a pre-determined extraction level.

This is the mechanism by which two flours can share an overall extraction rate and still differ in composition, because the mill can reach that extraction by combining different fractions. Roller milling is precise, it separates the germ cleanly (which is why refined roller-milled flour keeps so well), and it generates relatively little starch damage compared to stone grinding and is the industrial standard for these reasons.

Stone Milling

Stone milling is the older method, whereby rough natural or synthetic stones grind the entire kernel together in a single pass, crushing the endosperm across the germ and shearing the bran into fine particles. Nothing is separated out during grinding, so the natural result is whole-grain flour that needs to be sifted aggressively in progressively finer screen to achieve the extraction required. Stone-milled flour tends to carry more of the germ, and with it more flavor and aroma, but at the cost of a shorter shelf life due to increased fats subject to oxidation.

"Stone-ground" is not a synonym for "whole grain," and it is not automatically superior. A stone-milled flour requires more work to do so, but can easily be sifted to a refined white, and a whole-grain flour can be produced by roller milling and recombining all the fractions. And while the process of stone milling tends to take more skill and care by the miller than automated industrial roller mills, whether a flour was stone-milled is less indicative of the quality than how much of the grain remains, what was removed, how fine the particles are, and whether it was recombined after sifting.

Impact Milling

A third method shows up mostly in home mills and some specialty operations. Impact milling uses high-velocity rotor blades or pins to pulverize grain striking them in midair, shattering the kernel against the machine rather than crushing it between surfaces. It processes whole grain without the sustained friction heat of stones, and it excels at producing fine flour quickly, but it offers less tactile control than a stone mill and is less suited to coarse or cracked-grain work. But for a home baker who wants fine whole-grain flour and doesn't need infinite texture adjustment, it can be an economical and legitimate choice.

Damaged Starch

The damage the endosperm suffers in the milling process can help explain a flour's hydration behavior when no other number can. Two flours that match on protein, extraction, and even particle size, can still absorb water completely differently, because during milling, some starch granules get physically fractured, and broken granules absorb far more water than intact ones. Intact starch takes up roughly half its weight in water, while damaged starch can absorb up to twice its weight.

How much damage a flour suffers depends on the hardness of the wheat combined with the milling method. Hard wheats ground on steel rollers take more force than soft wheat and generate more damage (commonly 6–9%) than soft wheats (1–3%), while stone milling's shearing action also increases the chance of damage. A moderate amount is actually useful, boosting early water absorption and giving yeast an easier supply of fermentable sugar, but too much turns against you during a long fermentation because the amylases break down all that vulnerable starch, releasing the water it was holding back into the dough. This can cause a late stage failure in which a previously strong dough grows suddenly slack, sticky, and hard to shape as it ferments. This is one of the clearest reasons flour behavior can't be read from a spec sheet alone and why damaged starch percentage rarely appears on the label at all.

For the curious, damaged starch is measured in the lab either by enzymatic digestion or with an amperometric instrument like the SDmatic, which reads how much iodine the starch takes up. You will almost never see the figure on a retail bag, but mills track it closely, because it is one of the levers they pull to keep a flour behaving the same from lot to lot.

Flour Aging

Fresh flour and aged flour behave very differently, and knowing the difference can help solve a real bench problem: why a flour that felt slack and sticky last month handles better today. From the moment the kernel is broken, flour begins to age. And while freshly milled flour is more aromatic and flavorful, it's also weaker, stickier, and less predictable, because its gluten-forming proteins haven't yet oxidized. As flour ages, oxygen converts reactive sulfhydryl groups into disulfide bonds that reinforce the gluten network, strengthening the dough and lightening the color.

Natural aging takes several weeks, but commercial mills rarely wait that long, which is why aging is often accelerated chemically via controlled bleaching. The tradeoff cuts both ways: fresh flour usually tastes better, but aged flour usually behaves better. For whole-grain flour the process happens even faster and less forgivingly, because the germ's oils oxidize toward rancidity, so long room-temperature aging isn't a virtue there, it's a countdown.

If it helps to think in stages, flour moves through roughly five of them. Fresh-milled or "green" flour, used within hours or days, is aromatic and creamy-yellow but weak and unpredictable, its gluten proteins not yet oxidized. Fresh commercial flour has been milled, stabilized, and shipped within a few weeks and is the predictable industrial baseline. Naturally aged flour has rested three to six weeks with air, its sulfhydryl groups converting to strengthening disulfide bonds while the color lightens on its own. Artificially matured or bleached flour reaches that same place chemically in a fraction of the time, which is the subject of the next section. And old, oxidized flour has sat too long: the gluten degrades and the germ's oils turn, giving a rancid, cardboard smell and a grey, lifeless crumb.

Bleaching and Chemical Maturation

"Bleached" and "unbleached" are among the most misunderstood words on a flour bag. Freshly milled flour is faintly yellow and usually tastes better and behaves worse. Natural aging both whitens flour as it matures the gluten. Chemical treatment does the same, but on a predictable, industrial timetable instead of over weeks of exposure to oxygen.

Bleaching agents like benzoyl peroxide also strip yellow carotenoid pigments for a whiter flour without changing gluten behavior, while other maturing agents do the structural half of aging chemically, and chlorine gas, used for some cake flours, goes even further, lowering pH, weakening gluten, and modifying starch to absorb water fast, which is what makes ultra-tender high-ratio cakes possible.

Until recently, potassium bromate was used to maximize oven spring in industrial loaves, but it's a suspected carcinogen, banned across the EU, UK, Canada, and many other countries, and as of this writing banned in California as of 2027, with more U.S. states likely to follow. Most major American retail brands reformulated to unbromated years ago, though bromated flour still turns up in some commercial products.

Finally, a word on scary words vs scienctific facts. First, "unbleached" isn't automatically better: it only means the miller skipped chemical whitening, leaving them instead to either age the flour some other way or selling it younger and weaker. Second, bleached flour is not dead flour (there's no bleach left in the finished product), and it can start a sourdough culture perfectly well, despite rumors to the contrary. Personally, I don't bake with chemically bleached flour. I try to limit chemical processing and chemical exposure, even thoughs deemed safe, and I'd rather manage young flour myself anyway, but that's a preference. Your individual mileage may vary.

Malt and Enzyme Additives

When the Falling Number is high, meaning the weighted plunger took a long time to fall through a slurry left thick by low enzyme activity, mills often correct the insufficiency with a small fraction of low- or high-Lintner diastatic malted barley flour or fungal amylase, bringing activity into the useful range. Some home bakers do the same with commercially available diastatic malt additives when they want higher enzymatic activity and the advantages that come with it.

Diastatic malt is not to be confused with non-diastatic malt, and they do similar jobs achieved in very different ways. Diastatic malt is sprouted and dried gently to keep its enzymes alive so it can actively convert starch to sugar, speeding fermentation and deepening crust color through browning. Non-diastatic malt is roasted hot enough to kill enzymes so it's purely a sweetening, flavor and color ingredient, most evident in the sweet toasted note in bagels and malted-milk flavor. It has no enzymatic effect at all. The one additive I reach for myself is a little diastatic malt with certain low-extraction white flours, for browning and a livelier fermentation; it's also standard in bagels and pizza. As for non diastatic malt, well, I never drink a chocolate milkshake without it, and my childhood was made sweeter by chocolate covered malt balls, but that's a different story entirely.

Enrichment and Other Additives

The term enriched causes needless worry and deserves a plain explanation. U.S. regulations require refined white flour to have certain nutrients added back after milling, to make up for what refining strips out. Iron, several B vitamins, and folic acid are reintroduced simply to replace some of what's lost. Enrichment is a government-mandated public-health measure, and folic acid fortification in particular has had real benefits in populations that rely on refined grain. It has essentially no effect on flavor, water absorption, or how the dough behaves.

And if you're the sort who'd rather have the nutrition the grain came with than nutrition added artificially back at the mill, enriched flour is easy enough to avoid: you may prefer whole or fresh-milled grain anyway, since adding back a short list of vitamins raises the numbers but doesn't restore the flavor, the fuller nutrition, or the behavior of flour with all its bran, germ, oils, and fiber intact. Premium brands often leave their flour unenriched, while mass-market brands usually enrich by default, but you can't predict quality by this alone.

Commercial Specifications

A few other traditional milling terms describe how a flour was formulated from the mill's fraction streams, and they're useful to recognize because they can suggest behavior the protein number won't.

"Patent flour" is the highest-refined grade, drawn from the cleanest, lowest-ash streams of the inner endosperm — bright, predictable, low ash. "First clear" is the stream left after the patent flour is pulled off, high in protein but also high in ash, dark and strongly flavored, traditionally blended into rye breads to lend structure. "Artisan bread flour" is generally a marketing term for an unbleached, unbromated hard-wheat flour with a slightly elevated ash (around 0.5–0.6%) and moderate protein (roughly 11.5–12.5%), aimed at long sourdough fermentations. None of these names dictates a fixed extraction percentage, and they describe stream selection and miller's intent, not a guaranteed number.

International Labeling Systems

If you've ever been confused by a flour that says French T65 or Italian Tipo 00, the good news is that most national systems grade flour by ash content, how much of the mineral-rich outer grain remains, not by protein and not by fineness.

The French T-number is also based on ash content, but shifted by a decimal: T55 is about 0.55% ash (standard white), T65 about 0.65% (traditional artisan bread flour), T80 about 0.80% (high-extraction), on up to T150 for whole grain. Confusingly, T65 is not 65% extraction. The number is just ash, and higher simply means more of the outer grain. German type numbers (405, 550, 812, and so on) work the same way, the number tracking ash.

The Italian system causes the most confusion and can be a big pain in the ash to understand, so it's worth stating carefully. Tipo 00, 0, 1, 2, and integrale are defined by Italian law according to ash content, with 00 the most refined with the lowest ash and integrale, the whole grain. The persistent myth, especially in America, is that "00" means "finely milled." It does not. The classification is legally an ash standard, with no fineness requirement written into it. The myth does have factual basis, however, which is why it's so tricky: 00 flour is drawn from the soft inner endosperm, which naturally mills to a fine, powdery texture, and low-ash roller milling tends to produce fine particles anyway. So 00 flour usually is fine, but the fineness is an incidental side effect, not the definition, and you can have low-ash 00 at a coarser grind or fine flour that isn't 00 at all. And crucially, "00" tells you nothing about protein or strength: Tipo 00 flours range from soft, low-protein pastry flours to strong, high-protein pizza and bread flours. Two bags both marked 00 can behave nothing alike.

Part III — Buying Flour

Everything so far has been about what flour is. This part is about what you'll actually find when you go looking for it — how the market is organized, why the same category name means different things from different brands, and how to read the landscape so your choices are deliberate rather than accidental. Market details shift over time, so this describes the structure rather than chasing exact numbers.

Commercial Flour Market

Because corporate portfolios and distribution networks shift over time, this market overview captures the modern retail landscape, but the retail flour market sorts into a few recognizable tiers defined by volume dominance, corporate and brand positioning, and deep regional preference fragmentation. A choice of flour for biscuits is as brand loyal as a football team to many Southern households, and "if it was good enough for my grandma, it's good enough for me" has been heard in many a grocery aisle. In fact, it explains why self rising flour, a convenience product containing baking soda and powder, continues to sell, even though it's a one use product that is easily replicated at home.

Knowing which tier a flour comes from tells you what to expect from it. At the top by volume are the mass-market staples — Gold Medal and Pillsbury — the flours most home bakers grow up with and most likely to be on any given supermarket shelf. Alongside them, private-label store brands occupy the same tier at lower prices. In the premium national space, King Arthur has earned a commanding position among serious home bakers, built on tight quality control and strict protein specifications. And in specialty and alternative grains, Bob's Red Mill leads the mainstream supermarket presence, having introduced a generation of home bakers to whole and heritage grains.

I am quoting market-share percentages below, but I realize they shift year to year and can date a book quickly, so they are only here for illustrative purposes. The tiers themselves are stable, and the tier is what matters: it tells you whether you're buying a commodity blend built for consistency, a premium flour built to a tight spec, or a specialty product built for a particular grain.

All-Purpose Versus Bread Flour

Here's a fact that surprises people: "all-purpose" is a marketing category, not a technical specification, and is generally based on the flour being neither as soft as cake and pastry flour, nor as hard as bread or high protein bagel flour. There's no regulatory standard defining what hardness all-purpose flour must actually be, so its protein percentage, wheat base, and behavior vary dramatically from brand to brand. The name promises versatility, but guarantees nothing specific in this regard.

And the differences are not small. A premium all-purpose flour can be built from hard red winter wheat at nearly 12% protein and behave almost like a light bread flour. In fact, my experience with Costco AP flour, at 11.5 stated gluten protein, behaves exactly like Central Milling Baker's Craft Plus on bench and palate. A Southern all-purpose can be milled from soft red winter wheat at 9% protein and behave like a pastry flour, which may be attributable to the regional affinity for biscuits and other soft flour loving products. Both say "all-purpose" on the bag, but will produce completely different bread. This is why swapping one brand's all-purpose for another can quietly change your hydration, dough strength, and your results in ways completely unpredictable from the label nomenclature alone. The practical lesson is don't trust the category name, look at the protein, and if you can, the wheat class and milling style of the product rather than words or numbers on a bag.

"Bread flour" is a little more dependably harder wheat, since the name at least signals a higher-protein product appropriate for long fermentation naturally leavened products, but it still is set to no exact extraction or protein number by law or standard. Two bread flours can differ by a couple of points of protein and a wheat class, which is more than enough to change how they handle. See my test protocol and results for more on this phenomenon.

Regional and Artisan Mills

Beyond the national brands, the flour market fragments by region, shaped by region preferences, soil and weather patterns dictating which wheat grows nearby, and which and what type of mills serve the area. Recognizing the regional patterns helps you understand a flour before you've baked with it.

In the American South, brands like White Lily and Martha White are milled almost entirely from soft red winter wheat at low protein, around 8–9%, and often bleached. That weak gluten is a liability for hearth loaves and high-hydration sourdough but exactly right for the light, tender biscuits the region is known for. On the West Coast and Pacific Northwest, mills like Central Milling, Cairnspring, and Giusto's focus on regionally grown wheats, including hard whites and heirloom varieties, and lean toward stone milling and higher-extraction flours prized by artisan bakers for flavor and water absorption. Across the Intermountain West and Northern Plains, brands like Wheat Montana draw on high-protein hard red spring wheat grown at altitude under dry conditions, giving robust, tolerant flours suited to whole-grain and high-volume baking.

I promise there won't be a test and you don't need to memorize brands, but a flour's region can help tell you something about the wheat behind it, and the wheat behind it tells you how it will probably behave before you've opened the bag and just one more arrow in your informational quiver.

Organic, Conventional, and Identity Preservation

Three controversial and confusing supply-chain terms get tangled together on flour bags, and are well worth knowing a little more about.

"Organic" is a regulatory standard about agricultural inputs requiring growers use no synthetic fertilizers, no persistent chemical pesticides, no irradiation, and at the mill no chemical bleaching, bromating, or synthetic maturing for at least three years before getting to call themselves organic. What organic does not mean is single-origin, or even necessarily higher quality. Large organic mills routinely blend organic grain from many farms and regions to hit consistent commercial protein and other targets, so organic and traceable are different things, and if it is important you know the exact source of your wheatberries or flour, you'll need a different type of product.

Identity preservation is a separate, traceable quality, independent of organic status. An identity-preserved flour keeps a specific lot of grain isolated from harvest to bag, so it can be traced to a single farm, field, or cultivar — preserving the local character that's lost when grain is aggregated into regional silos.

Unlike organic, conventional cultivation sometimes uses glyphosate as a pre-harvest desiccant, though the practice varies widely by region, weather, and buyer, and is not universal; conventional storage may also use synthetic fungicides barred in organic supply chains. But for most bakers the more useful question isn't organic versus conventional so much as whether you want the traceable character of an identity-preserved flour or the consistency of a blended commercial one. Both can make excellent bread.

Warehouse and Value Sources

One practical note worth its own mention: warehouse stores can be an excellent and underrated source of good flour at low prices. Many artisan bakers buy from them, and some house brands are rumored to be private-label runs from well-regarded mills — a claim I have heard and mentioned in passing that's plausible but hard to confirm, and varies by region. The catch is consistency: warehouse flour offerings change by region, season, distributor, and even individual store, so a flour you love in one city may vanish or never appear in another. Still, they are worth watching for good value, but probably not worth building a formula around unless you've confirmed it's reliably stocked where you are.

Part IV — BreadLab

BreadLab is my attempt to record both qualitative and quantitative elements of flour used for making bread as I've tried to understand them in my own kitchen and experience. The method behind the observations you'll see tagged throughout this book are indicative of how I think about what actually matters when choosing a flour. This is the most personal part of the system I developed and use everyday, and also the most opinionated.

The BreadLab Method

BreadLab is not a laboratory and I am not running experiments in a strictly formal sense. It's a discipline: I bake the same reference formula over and over, changing only one (or as few) variables as possible, so that when a dough behaves differently I have a better chance of knowing why. The value isn't in any single bake — it's in doing enough of them the same way that patterns start to separate themselves from the noise over time.

Scientific research is meant to draw valid conclusions through controlled design, replication, and statistical analysis, but because I'm not running a real lab and can't bake the thousands of loaves under the highly controlled conditions required to establish that kind of rigor, BreadLab instead seeks practical understanding through standardized observation under consistent conditions. Those are different goals, and I'm doing the second, not the first. What I can offer isn't proof. It's careful, documented, repeatable observation, which is what good bakers have always relied on, and which I've simply tried to make more systematic and turn into a method you can follow to make your own discoveries.

A (Semi) Standardized Protocol

Two different questions you might ask about "how did you test these flours," and keeping them separate is what makes the answer meaningful: which flours did you test, and how. The how is the standardized protocol, and its whole purpose is to hold everything constant except the flour.

Every evaluation I did used the same reference formula, the same effective hydration as a starting point, the same inoculation and salt percent, the same mixing sequence, the same fermentation targets, the same shaping, the same baking environment, and the same evaluation criteria recorded the same way. When only the flour changes, differences in the dough can be attributed to the flour with some confidence, so when a flour struggles at the baseline hydration, only then do I adjust the hydration alone, so that I learn what each flour requires rather than guessing upfront what they might need based on the label alone.

That last point matters and it's worth stating as its own lesson: the baseline hydration is a diagnostic starting point, not a verdict. A flour that struggles at the reference number isn't a bad flour; it's a flour that requires different hydration number, and finding that number is the point of the exercise.

BreadLab Flour Library

The observations in this reference are drawn from my ongoing standardized evaluations, and I think it's only fair to tell you exactly which flours those are rather than gesturing vaguely at "my testing." As of this writing the library includes 365 Organic Bread Flour, Farmer Direct Organic Bread Flour, Bob's Red Mill Organic Artisan Bread Flour, Bob's Red Mill Organic All-Purpose Flour, Central Milling Baker's Craft, Central Milling High Mountain, Grand Moulins de Paris Tradition T65, and multiple freshly milled flour trials.

I know this is not a statistically representative sample of the flour market, and I don't present it as one. It's a deliberately chosen cross-section of national brands a reader is likely to find, plus premium benchmarks, regional and specialty mills, a French tradition flour, and fresh-milled flour as its own category. And testing is ongoing, so it grows steadily as I test more. When I cite a BreadLab Observation, it comes from this extremely time-consuming body of work, so if the list looks short to you, it's because it is: this is a record of careful observation, not a survey, and I'd rather tell you exactly what I've done than imply I've done more and I will continue to publish more results as they become available.

BreadLab Observations

A handful of patterns have come up often enough across my evaluations that I've come to trust them, while still holding them as observations rather than laws. The most important one runs through this entire book: across my flour evaluations, I've repeatedly found that published protein percentage alone is a poor predictor of how a dough will behave on the bench or in the oven or on the plate. Flours with similar protein numbers routinely handle differently, and a spec sheet is a starting point rather than a forecast.

A second pattern: the flours I found easiest and most rewarding to work with weren't always the ones with the most impressive specifications or the most artisan pedigree. Availability and handling have surprised me more than once, with widely-available supermarket flours outperforming my expectations and boutique flours demanding more management than their reputation suggested. I'll name no winners here, because the point isn't a ranking — it's that the bench keeps refusing to honor the hierarchy the labels imply.

In any event, the rating is subjective and in large measure based on what I think will be an appropriate balance of difficulty to handle and finished product quality for any given experience and skill level. What I prefer to make for my friends and family is based on my unique experience and skill level. You may require something completely different, and I've tried to keep the two separate and clear in my mind and in this book.

Working Hypotheses

An observation tells you what happened. It doesn't tell you why, and I want to keep those two things clearly separate, because my explanations are less settled than my observations. What follows is my current thinking, offered as hypothesis, not conclusion.

My current working hypothesis for why protein percentage predicts so poorly is that the handling differences between flours of similar protein are driven mostly by the interaction of two things the number doesn't capture: the milling method, which governs damaged starch and bran particle size, and the glutenin-to-gliadin ratio of the specific cultivar, which governs whether the protein that's present can actually build strong gluten. I also suspect that high protein numbers in some stone-ground whole-grain flours are partly inflated by non-gluten proteins in the bran and germ, which would make the number look stronger than the dough behaves.

I hold these loosely. If further evaluation contradicts them, I'll revise them, and that's exactly how this is supposed to work: observe, explain provisionally, keep testing, update when the evidence warrants. A working hypothesis that turns out wrong hasn't damaged anything — it's modeled how a careful baker should think.

What Matters Most

People want a single number to rank flours by, and the number they reach for is almost always protein. If I had to rank the flour characteristics that most influence my own sourdough baking, protein percentage would sit well down the list, but not because it's meaningless, but because so many other things matter more to me and, I believe, to you, my reader. This ordering is editorial: it's how I weight these factors after baking a great many loaves, offered as judgment, not as measurement. In fact, it's the most prominent and important opinion in the opinionated baker, and I take my recommendations very, very seriously.

When I pick a flour blend for my own bread, I consider a lot of factors, and protein level is important in predicting how tough or tender the crumb will be, a major element to control if you know you want a sturdy sandwhich bread or a tender appetizer bread.

But for the book, I had to consider a lot of other factors I think are important when starting this journey. So follows is a list, roughly in order of how much they steer my results are overall handling experiences like behavior and strength on the bench, which is the thing all the numbers are trying and often failing to predict; flour type and extraction, which set water capacity (hydration) and baseline character; protein quality, meaning the glutenin-to-gliadin balance rather than the raw amount; enzyme activity; damaged starch; and only then protein percentage, useful as a bulk clue but secondary to all of the above. Below those, ash content, milling method, wheat variety, and the flour's age and freshness each contribute their part.

If there's one habit this whole reference is trying to build, it's this: stop only asking what a flour's protein number is and start also asking what else the flour actually does. The protein number is where the conversation starts, but it was never supposed to be where it ends.