Sourdough starter with a whole-wheat first feed

A starter in its making

This recipe records the beginning of a sourdough starter, not a finished loaf and not a packaged culture with known strains. Its first mixture is equal weights of whole-wheat flour and cool water: 50 grams of each. The paste sits loosely covered for a day, then a measured portion is refreshed with all-purpose flour and water. Daily volume is kept small by retaining 50 grams and discarding the rest; after the first refresh, the routine becomes twice daily. The starter is a flour-and-water culture being established by repeated additions of fresh food.

The title describes the particular sequence that makes this method distinctive: whole-wheat flour starts the culture, while all-purpose flour is used for subsequent feeds. The method comes from King Arthur Baking's test kitchen, which reduced its earlier larger batch after baker feedback in February 2026. That date is the history of this published formula, not the invention of sourdough or a claim about the first use of a starter. Other bakers build and maintain starters with different flours and ratios.

Once established, the starter is used to leaven bread dough. A young culture is not ready simply because the calendar says day five or day seven. This method describes readiness as reliable doubling about six to eight hours after a feed. The time gives a practical reference point for this formulation and room-temperature practice; it is not a universal speed limit for every kitchen, flour, or culture. The recipe's small batch is intended to build a culture for ongoing baking, rather than to yield a fixed quantity of bread.

Why whole-wheat flour begins the culture

Whole-wheat flour brings the whole milled grain, including bran and germ as well as the starchy endosperm. King Arthur recommends beginning with whole-grain flour because it expects a richer field of naturally occurring organisms than in refined flour. That is a useful practical rationale for choosing the first feed, not a guarantee that every bag carries the same microbial population or that an individual starter will mature on the same timetable.

The flour also changes the mixture's physical character. Bran particles absorb water and remain visible, so a whole-wheat starter can look darker and more textured than one made with sifted white flour. In this recipe, the first mixture is a 1:1 flour-to-water ratio by weight. Later feeds use the same weight of flour and water, but the flour changes to all-purpose. The ingredients are kept simple so the reader can see how flour-and-water fermentation proceeds without fruit, dairy, sugar, or a commercial culture added.

Controlled research has found that flour type and extraction rate influence sourdough communities and their fermentation products. A study comparing refined and wholemeal soft and durum wheat flours reported differences in microbial composition and measured acids, sugars, and amino acids. Those results explain why flour is not an inert filler, but they do not identify the organisms in this home mixture or predict its flavor. The flour variety, milling, water, temperature, vessel, and the organisms that establish themselves together all matter.

A community grows through refreshment

During sourdough fermentation, yeasts produce carbon dioxide that can leaven a later dough, while lactic acid bacteria produce organic acids and other compounds that affect acidity and aroma. Sourdough starters commonly hold a community of yeasts and lactic acid bacteria, but their members and proportions vary. This recipe does not inoculate the flour with named species, so it cannot promise a specific organism or a fixed flavor profile. The microbial community is an outcome of the flour, environment, time, and repeated feeding.

Each refresh is a form of backslopping: some active mixture is carried into a new portion of flour and water. Keeping 50 grams while replacing the rest with fresh feed changes the balance between the existing culture and new substrate. Repeating the routine provides fresh flour and water while limiting how much starter accumulates. That is why the discard is part of the small-batch schedule rather than a mistake or a sign that the previous mixture failed.

Food scientists distinguish spontaneous flour-water starters from processes begun with selected commercial cultures, and from methods that combine a culture with later backslopping. This recipe begins spontaneously and then propagates the portion retained at each feed. Research on starter microbiomes has linked differences among cultures with measurable differences in bread acidity, volume, and crumb properties, but those studies do not let a baker infer exact results from a jar's appearance alone. A home starter's behavior belongs to its own conditions and feeding history.

The schedule and its changing rhythm

On the first day, the flour and water are mixed until no dry flour remains, loosely covered, and left at about 70°F (21°C) for 24 hours. This gives the freshly mixed flour time to hydrate and ferment before it is refreshed. On day two, 50 grams of the mixture are kept and combined with 50 grams water and 50 grams all-purpose flour. The retained portion carries the developing culture forward; the fresh feed supplies new flour and water without allowing the total mass to grow rapidly.

On day three, the recipe moves to two feeds, roughly twelve hours apart. At each feeding, retain 50 grams of starter and add 50 grams water plus 50 grams all-purpose flour. The same rhythm continues on days four and five, and days six and seven if the mixture is not rising reliably. By mass, each feed is one part retained starter, one part water, and one part flour. This 1:1:1 pattern is specific to the recipe's small batch; other feeding ratios are documented and produce different dilution and timing.

The timetable is a workable routine, but temperature changes the rate of fermentation. King Arthur advises cool water in a warm room and lukewarm water in a cool room, and explains that a colder environment slows starter growth. A twelve-hour interval therefore describes the intended cadence, not a promise that every starter reaches the same point at the same hour. If the culture is not consistently doubling after a feed, this recipe says to continue the twice-daily refreshes rather than declare it mature by date alone.

Equal weights of flour and water make the initial paste a 100% hydration mixture in baker's terms: water weighs as much as flour. The retained starter also contains flour and water, so the first refresh is not simply a fresh 50-gram portion of each. Keeping the recipe in grams makes those proportions repeatable and avoids the volume changes caused by flour being packed loosely or tightly into a cup. The small-batch weights are deliberately modest, but a sufficiently large jar is still useful because an active starter expands.

King Arthur notes that on day two a new mixture may show no obvious activity, or may already show bubbles or some growth. Either outcome can occur before the regular twice-daily rhythm begins. A single visible rise is therefore not the criterion for calling the culture established. The repeatable post-feed doubling cue later in the method gives the baker a more useful comparison across several feeds than a one-time burst of gas.

Equipment and signs of progress

Use a roomy, non-reactive vessel that leaves headspace for expansion. Glass, crockery, stainless steel, and food-grade plastic are suitable container materials in King Arthur's starter guidance. A loose cover keeps the mixture from drying while allowing pressure to escape as gas forms; this is not a sealed canning process. Mixing thoroughly at the start prevents dry pockets, and scraping the sides keeps old flour from building up above the active mixture.

Observe how the starter rises and falls between feeds. Bubbles, expansion, and a repeatable rise after feeding are useful process observations. The recipe uses consistent doubling six to eight hours after a feed as its readiness cue. A ruler, elastic band, or marked jar can make the change in height easier to compare, but the mark measures volume change, not microbial safety. A sluggish rise may reflect a cool room or a culture that needs more time; a single early burst is not the same as dependable activity across successive feeds.

Temperature changes how quickly the same ingredients move through the schedule. At a cooler room temperature, the starter may need longer to show the same rise; in a warmer room, the timing may compress. King Arthur's advice to choose cool or lukewarm water in response to the room is a way to moderate the starting conditions, not a substitute for observing the mixture. The recipe does not call for a laboratory incubator or a tightly controlled fermentation chamber.

Whole-grain flour can make the early mixture look different from the later all-purpose feeds: flecks of bran remain visible, and the paste can feel thicker as particles absorb water. Those appearances are expected consequences of using a less-refined flour, not a calibrated test of microbial activity. Record the feed time and starting height if the culture's pace is unclear; comparing it with its own previous feeds is more useful than comparing it with another baker's warmer kitchen.

History without an invented origin

Flour-and-water sourdough fermentation is old, but the origins of a particular household starter are rarely traceable. A food-science review describes sourdough as an ancient cereal fermentation and surveys how starters are propagated, while noting that flour, inoculum, and processing choices create different practices. Such evidence supports a long history of fermented cereal foods; it does not establish one birthplace, one first starter, or an uninterrupted line from an ancient bakery to a modern jar.

The recipe here has a much narrower documented history. King Arthur Baking's test kitchen states that its home starter method was updated in February 2026 to make a smaller amount in response to baker feedback. It is a contemporary American baking instruction, with metric weights and a domestic schedule, and its own source identifies the steps and quantities. Keeping those records separate avoids turning a modern measured recipe into an unsupported claim of cultural antiquity.

Variations and what this starter is for

King Arthur's broader guide offers pumpernickel flour as another whole-grain starting choice. This entry follows the whole-wheat route and changes to all-purpose flour for later feeds. Other documented starter practices vary flour, hydration, retained amount, feeding intervals, and whether a baker begins with spontaneous fermentation or a selected culture. These are not interchangeable adjustments: changing the flour or refresh ratio changes the mixture's composition and the amount of established culture carried into each feed.

The purpose of this starter is to raise bread dough. Once its response to feeding is reliable, a baker can use it in sourdough bread formulas and continue refreshing the portion kept for future baking. The starter itself is not the dough and does not determine a loaf's formula by itself; bread recipes differ in how much starter they use and how long they ferment. Its acidity and gas production develop during maintenance, while mixing, shaping, proofing, and baking happen later in a separate recipe.

Starter culture is not the same thing as the bread dough that it raises. In this build, flour and water are refreshed over several days until the culture responds consistently; later, a bread formula combines a portion of that starter with its own flour, water, salt, and fermentation schedule. The starter's acidity and gas production can influence a loaf, but the baker still controls the dough's hydration, mixing, shaping, and proof. A jar that is vigorous at one feeding does not replace the later dough's own timing.

Whole-wheat flour is the choice for this recipe's first mixture, while pumpernickel is a documented alternative in the source's wider guide. Some bakers keep a culture on whole-grain flour; others use refined flour for routine refreshes, as this recipe does after day one. Flour choice may alter texture, aroma, and fermentation behavior, but the cited comparison study was conducted under its own defined conditions. It should not be treated as a promise that switching flours will fix a sluggish jar or reproduce a particular loaf.

Food safety and practical limits

Raw flour is not ready-to-eat food and can carry pathogens. The CDC advises against tasting raw dough or batter and recommends cooking flour mixtures; wash hands and surfaces after handling raw flour. A starter is therefore a preparation ingredient, not something to taste raw as a safety test. The fact that a mixture has bubbles, a pleasant smell, or a high rise does not establish that raw flour is safe to eat.

This recipe describes the initial build and a readiness cue, but it does not provide a validated preservation process, a room-temperature shelf life, or storage limits for every established starter. Do not infer those details from the fermentation schedule. Follow a tested maintenance or storage method for the culture you keep, and bake dough thoroughly before eating. The measurements, 70°F starting environment, and six-to-eight-hour doubling cue belong to this particular King Arthur method; they are not food-safety critical limits.