PProvenance
Enter
P
Japanese · Nagasaki

Castella — The Portuguese Sponge That Became Japanese

Castella's technique is a study in controlled simplicity. The ingredients are four: eggs (in large quantity — typically 10 eggs per loaf), sugar (wasanbon or mizuame — Japanese refined sugar and starch syrup — for the finest versions), flour (low-protein cake flour), and honey (for flavour and for the characteristic stickiness of the base). There is no butter, no oil, no chemical leavening. The rise is entirely from egg foam. The batter is mixed differently from a génoise: rather than folding flour gently to preserve air, castella batter is stirred — the gluten development from stirring produces the slightly denser, more uniform crumb that distinguishes castella from génoise. This is the Japanese modification: where French technique seeks maximum aeration, castella technique seeks a specific density — a tightly textured, moist, almost custardy crumb.

‹Swipe for the next dish›
ServesOne loaf (approximately eight to ten slices)
Sourced2of 5 in BC
Cook
I

The Market List

The benchmark for each ingredient, and who stocks it near you. Tap a stockist to go straight to the product.

10
eggs
whole, at room temperature
Your own market
300 g
wasanbon
Japanese refined cane sugar; Western caster sugar is an inferior substitute
30 g
mizuame
thick starch syrup
Not yet stocked in British Columbia
250 g
cake flour
low-protein; sifted
Your own market
40 g
honey

Each tap is counted, so the stockist sees the demand Provenance sends them.

II

The Method

1
Beat the eggs with the wasanbon until the mixture is pale, voluminous, and holds a ribbon that dissolves slowly back into itself. The foam is the sole leavening agent; it must be fully developed before any other ingredient is added. The test: the ribbon holds on the surface for three to four seconds before sinking.
2
Add the honey and mizuame to the egg foam and incorporate evenly.
3
Add the sifted cake flour and stir the batter with a wooden paddle, always in the same direction. Do not fold. The stirring is deliberate: it develops minimal but consistent gluten, producing the tightly textured, moist crumb that distinguishes castella from génoise. This is where the dish lives or dies — too little stirring leaves an uneven crumb; too much toughens it. The batter is ready when it is smooth, slightly elastic, and pulls away from the paddle in a slow, continuous sheet.
4
Pour the batter into a wooden box mould lined with paper. The wooden construction insulates the batter during baking, moderating heat transfer and ensuring the even surface colour that defines the finished loaf. Bake until the top is a deep, even golden-brown with no pale patches and no dark spots.
5
Remove the castella from the oven and invert it immediately onto a board. Place a second weighted board on top and press overnight. The pressing consolidates the crumb into the dense, fine, uniform matrix of a correctly made Nagasaki castella, and produces the smooth, flat surfaces that define the finished product. The test: when the pressed loaf is cut, the cross-section shows fine, uniform, visible cells — not the open, irregular structure of génoise.
6
Touch the base of the finished loaf lightly with the tongue. A correctly made castella adheres slightly — a faint stickiness from the mizuame. This is the final test that distinguishes a properly made Nagasaki castella from every imitation.
III

Quality Hierarchy

01
Reserve — Wasanbon from Shikoku or Tokushima, produced by the traditional multi-day hand-pressing process; mizuame of the highest clarity; single-origin honey of pronounced character. A wooden box mould used throughout.
02
Estate — Wasanbon sourced from a named producer; good-quality mizuame; a clean, fragrant honey. Wooden box mould used.
03
Market — Wasanbon where obtainable, or high-grade Japanese caster sugar; mizuame from a reliable importer; a neutral-flavoured honey.
04
House — Western refined caster sugar in place of wasanbon; glucose syrup as an approximation for mizuame; commercial honey. The crumb will be coarser and the characteristic fineness of sweetness will be lost.
IV

Sensory Tests

◇Sight. The top surface is a deep, even golden-brown — no pale patches, no dark spots; consistent colour is evidence of correct batter temperature and wooden-box insulation.
◇Sight. The cut cross-section shows fine, uniform, visible cells — the dense, even matrix of castella, not the open, irregular structure of génoise.
◇Touch. The cut surface, pressed lightly, springs back slowly, holding the impression for two to three seconds before recovering.
◇Touch. The base of the finished loaf touched to the lip adheres slightly — the faint stickiness of mizuame, the defining mark of a correctly made Nagasaki castella.
◇Where it lives or dies. The batter must be stirred — not folded — with a wooden paddle, always in the same direction, until smooth and slightly elastic: under-stirring yields an uneven crumb, over-stirring toughens it, and neither produces the tightly textured, moist, almost custardy density that defines castella.
Pillar V
Cross-cuisine parallels are part of membership.
The dishes this one teaches you, across the cuisines — on every recipe.
Enter · from $49 a year
VI

What to Pour

The pairing
Nikka From The Barrel
Nikka Whisky — Miyagikyo Distillery · Miyagikyo — Nikka Distillery
Not yet stocked in British Columbia
Whisky
From British Columbia

Nothing from British Columbia holds against this yet.

Without alcohol

No pour without alcohol holds against this yet.

Why each one works is part of membership — Enter · from $49 a year →

The whole pairing
Pillar VII
Origin and lineage are part of membership.
Where the dish comes from and how it came to be — on every recipe.
Enter · from $49 a year