This Japanese-style milk bread uses tangzhong, a cooked paste made from part of the flour and milk, to help produce a soft crumb. The formula contains 400g total flour and makes about 788g dough before cooking losses. Shape three rolled logs, proof in an open 9-by-5-inch pan, and bake at 350F until set. Shokupan can also be square and lidded, but that requires a pan-specific dough quantity and proof below the lid. Tangzhong can improve softness; it does not guarantee a fixed number of fresh or mold-free days.
A soft milk loaf is useful for toast, sandwiches and slices that pull apart along the rolled sections. Here the dough contains milk, egg, butter and a small cooked flour paste. Those ingredients work together with mixing, fermentation and baking; there is no single ingredient that guarantees the softest bread.
The formula below retains this page’s original ingredient weights, with corrected arithmetic and a clearer open-pan method. Its exact pan fit, rise schedule and keeping quality still need a documented kitchen test. For a separately developed formula, follow King Arthur’s Japanese Milk Bread in full; its quantities differ. Allow a substantial part of the day, including cooling, rather than treating four hours as a deadline.
What Makes It Shokupan
Shokupan is a Japanese loaf-bread category, not a promise that every loaf contains egg or uses one particular cooked-paste method. Yamazaki’s bread encyclopedia distinguishes square loaves baked with a lid from domed, lidless yamagata loaves. Both are shokupan. Pasco’s ingredient FAQ identifies its Chojuku series as made without egg, a useful counterexample to an egg requirement. That ingredient statement is not a universal allergen or cross-contact assurance; check the current package.
Three rolled pieces are one shaping choice. They are logs, not braided strands. Other recipes use four pieces or a different arrangement. Folding and rolling can give distinct sections to pull apart, but the number three is not what defines Japanese milk bread.
What the Cooked Paste Changes
Cooking a portion of flour with liquid swells and gelatinizes some starch before the final dough is mixed. In this formula the paste starts at one part flour to five parts milk, and contains 6.25% of the total flour. Its thickening changes how the dough handles water. It does not mean every starch granule bursts simultaneously at exactly 65°C, or that all the held water becomes steam in the oven.
Tangzhong and yudane are related approaches, but their processes and ratios differ. A 2014 original yudane study used boiling water and flour at 1:1, with much larger treated-flour portions than this loaf. Those breads were initially softer and firmed more slowly, but also had lower specific loaf volumes under the tested conditions. The study supports a formulation-dependent effect, not a guarantee that this paste doubles shelf life or always increases rise.
For more background, see the tangzhong method guide. For the saucepan, the useful cue is a smooth, visibly thickened paste that leaves whisk tracks. A thermometer can help you observe the process; one exact reading is not a pass/fail test of all starch gelatinization.
The Formula
Yield: one loaf in an open 9-by-5-inch pan. This is not an interchangeable instruction for every 9-by-4-inch Pullman pan, whose depth and lid clearance matter.
Tangzhong
| Ingredient | Weight | Percentage of total flour |
|---|---|---|
| Bread flour | 25g | 6.25% |
| Whole milk | 125g | 31.25% |
Whisk together in a small saucepan. Cook gently, stirring continuously, until thickened and leaving tracks. Expect a few minutes, with the actual time depending on the pan and heat. Transfer to a shallow container and cool to lukewarm before mixing. The starting weight is 150g; evaporation means the finished paste can weigh less. Use all of it rather than adding a second 150g portion.
For advance preparation, cool promptly, cover and refrigerate at 40°F/4°C or below; use the next day. Do not leave the milk-and-flour paste out overnight. Use a suitable covered container rather than pressing unsuitable plastic onto a hot saucepan.
Main dough
All percentages use 400g total flour, including the 25g already in the paste.
| Ingredient | Weight | Baker’s percentage |
|---|---|---|
| Bread flour | 375g | 93.75% |
| Whole milk | 120g | 30% |
| Cooled tangzhong | All of the batch above | Already counted |
| Beaten egg, without shell | 50g | 12.5% |
| Sugar | 40g | 10% |
| Salt | 8g | 2% |
| Instant dry yeast | 5g | 1.25% |
| Unsalted butter, softened | 40g | 10% |
The combined ingredient mass is 788g before roux evaporation and handling losses, excluding wash and pan grease. Weigh the finished dough before dividing it.
Milk plus egg weighs 295g, or 73.75% of flour weight. That is an ingredient ratio, not pure-water hydration. Using approximate water contents of 87% for whole milk, 74% for egg and 16% for butter gives about 64.1% estimated water, before evaporation. Excluding butter’s water gives about 62.5%. Products vary; neither calculation supports the old 67% claim. King Arthur’s hydration explanation sets out these approximate water fractions.
Try It Yourself
The Baker's Bench
Explore representative bread formulas in the Baker's Bench. For this loaf, keep the 400g total-flour basis and both tangzhong ingredients in your own scaling calculation; a preset does not establish the right loading for every Pullman pan.
Open ToolEquipment
Use a scale, small saucepan, whisk, mixing bowl, probe thermometer, rack and an open 9-by-5-inch loaf pan. A dough-hook mixer can help, but hand kneading is possible. Follow the machine’s capacity, speed and run-time limits.
Method
1. Mix and develop the dough
Make and cool the paste first. Combine the main flour, sugar, yeast and salt. Add the main milk, egg and cooled paste and mix until evenly moistened. Knead until the dough is cohesive and beginning to stretch.
Do not automatically heat the milk to 90–95°F: the paste, room and mixer also affect final dough temperature. A moderately warm dough is easier to schedule than an overheated one. Desired dough temperature depends on the actual ingredients and mixing process.
Add the softened butter in several portions, working each in before adding more. Continue until smooth and elastic, with a thin membrane when gently stretched. The windowpane test helps assess development; do not chase a fixed number of minutes after the dough is ready.
Use the mixer manufacturer’s dough setting. KitchenAid specifies Speed 2 for yeast dough, so generic instructions to knead on medium or medium-high are inappropriate.
Delayed butter addition is useful in many brioche methods, but early fat does not make gluten development impossible. Published milk-bread formulas also use an all-in method with melted butter. Keep the softened-butter sequence here without turning it into a universal rule.
2. Bulk ferment
Cover in a lightly greased bowl and let the dough become substantially expanded and aerated. The inherited estimate is 1.5–2 hours around 75–78°F; cooler dough or a different yeast can take longer. One gentle fold during the early rise can help if the dough needs strength.
Sugar is not simply an accelerator. At higher concentrations it can slow fermentation through osmotic effects, with yeast strains responding differently. The 10% sugar in this dough is no reason to assume a rapid rise. Watch the dough rather than declaring it ready at two hours.
3. Shape three sections
Weigh and divide into three equal pieces, nominally about 263g each before losses. Flatten each into a rectangle, fold it in thirds and roll it into a log sized to fit across the pan. If it resists stretching, cover and let it relax briefly rather than tearing it.
Place the logs seam-down, side by side along the length of the greased pan. Aim for even pieces and a secure seam, not maximum tension or forced compression.
4. Proof in the open pan
Cover and let the loaf become puffy and well expanded. The original 45–60 minutes is a starting estimate. A crown around an inch above the rim can be useful in this open pan, but flour strength, actual pan dimensions and shaping affect that height. Use overall expansion and dough feel too.
A light poke test can support the assessment; no single dent response is conclusive for every enriched dough. A collapsing loaf may already be overproofed and can bake with reduced volume.
For a lidded Pullman loaf, use a recipe specified for that pan and leave the required clearance below its lid. Do not proof an inch above the rim and then force the lid shut. An open loaf’s height cue cannot be transferred to a closed pan.
5. Bake and cool
Preheat to 350°F. If desired, lightly brush with an egg beaten with a tablespoon of milk; you will not need all the wash. Discard leftover raw-egg wash after use.
Begin checking around 30–35 minutes. Look for a set loaf and golden top; a center reading around 190–195°F is a supporting cue. Continue as needed and tent loosely with foil if the top colors too quickly. Oven calibration and pan material affect the result. There is no professional rule that all enriched breads bake at 380°F, nor a rule that every milk bread must stay at or below 350°F.
Let the loaf settle briefly, then use dry mitts to turn it out safely onto a rack. Avoid leaving it to cool completely in the pan, where moisture can collect. Cool fully before slicing and bagging; hot slicing can make even a baked loaf seem gummy.
What Tangzhong Adds to a White Loaf
Effects depend on the full formula and baking process
| Feature | 01Without a cooked paste | 02With tangzhong |
|---|---|---|
| Texture | Can be soft and springy | Can improve softness in a suitable formula |
| Water handling | Depends on flour and enrichments | Pre-cooked starch changes water handling |
| Keeping quality | Depends on recipe and storage | May stay softer longer; no fixed day guarantee |
| Staling | Includes starch and moisture changes | Still occurs; the gel does not immobilize all water |
| Crumb | Affected by mixing, proof and shaping | Still affected by mixing, proof and shaping |
| Preparation | No paste to cook | Cook and cool a portion of flour and liquid |
This is a method comparison, not a controlled test of two loaves at matched hydration.
jayarrbread.comSoftness, Variations and Storage
Ordinary sandwich bread need not be dry. Milk, sugar, egg and butter change crumb and crust as well as flavor; their effects cannot be separated from tangzhong without a controlled comparison. A gel does not prevent all starch recrystallization or stop moisture movement.
For rolls, twelve equal pieces would be about 66g before losses, not 68g. Their proofing, pan fit and baking time need checking as a separate format; touching neighboring rolls is not, by itself, proof that they are ready. See the dinner-roll guide for more on shaping and proofing.
Chocolate or cinnamon fillings change the dough’s bake and storage needs. Custard and meat-filled buns especially need a complete recipe covering filling preparation, cooking and refrigeration. The plain-loaf method does not validate a universal 15–18-minute bake for stuffed buns or safe storage at room temperature.
Dairy-free adaptation: replace milk in both the paste and final dough, replace the butter, and omit or replace the milk in the wash. Replacing milk alone leaves dairy in the loaf. Eggs are not dairy, so this would still not be vegan. For milk allergy, check every ingredient and cross-contact information. King Arthur’s dairy-free baking trials include milk-bread rolls made with plant milk and vegan butter, but do not establish that all substitutes behave identically in this formula.
Cool the plain loaf completely and wrap it to limit drying. Use it promptly at a cool room temperature or freeze slices for longer keeping; two or three months is a practical freezer-quality target, not a sharp safety deadline. Toast slices from frozen if desired.
Refrigeration often speeds bread firming, but it is not forbidden: hot, humid conditions or perishable fillings can make refrigeration appropriate. King Arthur’s storage trials discuss the tradeoff. Tangzhong is not a preservative, and no four- or five-day mold-free period is assured. Discard a moldy loaf rather than trimming it. Our bread-storage guide covers the distinction between staling, drying and spoilage.
Troubleshooting
Dense loaf: check yeast activity, dough development and proofing. Underproofing is one possibility; overproofing, weak structure or a poorly matched pan can also matter. Failure to reach one inch above a particular pan is not a universal diagnosis.
Brown top, unfinished center: tent and continue baking. Check the oven temperature and pan position. The crust contains browning reactants and dries sooner than the center, so crust color alone does not establish doneness.
Dough seems broken after butter: allow time for gradual incorporation, within the mixer’s limits. Very cold butter disperses poorly; excessive mixing and heat can also damage the result. Rest briefly if needed rather than increasing speed indiscriminately.
Dry after a couple of days: examine wrapping, bake duration, flour measurement and overall formula before blaming the paste. There is no exact 65°C switch below which the paste has no effect.
Gummy base: insufficient baking, slicing warm, proofing issues or condensation may contribute. Cool on a rack once safe to handle; immediate, hurried unmolding is not a cure for every gummy crumb.
Frequently Asked Questions
- What is Japanese milk bread or shokupan?
- Shokupan is Japanese loaf bread and includes square lidded and domed lidless forms. This milk-and-egg version uses tangzhong and three rolled sections, but neither egg nor this precise method is a requirement for every shokupan.
- What is tangzhong and why can it make bread softer?
- It is a cooked paste made from part of the flour and liquid. Pre-cooking changes starch and water handling and can improve softness in a suitable formula. It does not guarantee a larger loaf, prevent all staling or establish a fixed shelf life.
- Can I use water or plant milk instead of dairy milk?
- A flour paste can be made with water or plant milk, but the loaf's composition and texture will change. A dairy-free version must replace both milk portions, the butter and any dairy in the wash. Eggs can remain in a dairy-free loaf, so dairy-free does not mean vegan. Check labels and cross-contact information for milk allergy.
- Why does milk bread brown before the middle is done?
- The surface dries and heats while the center stays moist. Milk solids, egg and sugar affect browning. For this open loaf, use the stated 350F starting temperature, tent if needed, and check that the center is baked rather than judging only by top color. Pan and oven differences affect time.
- How does milk bread differ from brioche?
- Recipes overlap, but brioche often contains more butter and egg than this loaf, which has butter at 10% of flour weight. Butter at 50% means half the flour weight, not equal parts. Either bread can have a light crumb when well made; softness is not explained by fat quantity or tangzhong alone.
