Vegan French Toast Without Egg or Banana

by
Three thick slices of french toast stacked and cut through, the cut faces showing a pale matte interior between two mottled bronze surfaces

Vegan french toast fails in one of two ways, and most people get both on the same plate. The middle stays wet, so the slice tears when you cut it. The surface never colors, so it comes out pale and faintly rubbery instead of browned. Neither problem is caused by leaving the egg out. They are caused by not replacing the specific things the egg was doing.

Nine tested recipes sit behind this page, plus four peer reviewed papers, two manufacturer product pages and the food safety guidance from FDA and Nebraska Extension. The developers disagree with each other about soak time, about pan heat and about whether stale bread is better than fresh, and where they disagree this page says so and names who published what. No banana anywhere. The reason for that is further down, and it is a taste problem rather than a technique problem.

The three jobs the egg was doing

A 2022 review of vegan egg ingredients in the journal Foods states the functions plainly: “Eggs play significant roles in foods, owing to its gelling, foaming, and emulsification features.” For french toast the one that matters is gelling. The custard has to thicken enough to cling to the slice, then set inside the bread once it hits the pan. If it never thickens, it runs off. If it never sets, you get a wet middle.

Browning is a separate reaction and the same review does not cover it. A 2026 review in Toxics describes the mechanism: the Maillard reaction “includes the interaction of reducing sugars and amino acids at high temperatures, leading to browning.” Sugar alone will not do it. There has to be something in the custard carrying amino acids, which in practice means a plant milk with real protein in it, or tofu, or a legume flour.

The same review adds a detail worth keeping: “crust development, water loss, and surface protein denaturation, along with process time and temperature, were driving factors.” Water has to leave the surface before color arrives. A slice dripping with custard steams itself and stays pale no matter how long you leave it there.

That gives you the replacement list. A thickener that also sets under heat, which in practice means starch, legume flour or blended tofu. A liquid carrying enough protein and fat to brown, which is the plant milk decision most recipes skip past in half a sentence. And a pan hot enough to dry the surface before it colors, which is the last section on this page.

Building the custard without banana

Starch is the standard answer, and it is not a home cook workaround. The Foods review notes that “Starches from pulses are also increasingly used in formulating vegan eggs to play the role of binding and thickening.” Every tested developer in this set except one reaches for a starch, a legume flour, tofu or chia.

There are four routes and they behave differently in the bowl and in the pan.

Starch and plant milk. Sam Turnbull uses 1 cup of plant milk, 2 tablespoons of cornstarch and 2 tablespoons of all purpose flour for four to six slices. Nora Taylor runs a much higher starch load, a quarter cup of cornstarch to 1 cup of soy milk, plus a teaspoon of ground flaxseed and half a teaspoon of baking powder. Bob’s Red Mill describes cornstarch as “a dense powder made from the endosperm portion of the corn kernel” sold for thickening. It is the easy vegan french toast route, in the sense that everything in it is already in the cupboard, and it needs one piece of maintenance: Taylor tells you to whisk the batter again right before each slice, “as the cornstarch will settle to the bottom of the bowl.” She is right, and skipping it is why the last two slices come out thin.

Chickpea flour and plant milk. Alison Andrews uses 1 cup of soy milk, 4 tablespoons of chickpea flour and 2 tablespoons of nutritional yeast for six thick slices. Chickpea flour is not a starch. Bob’s Red Mill sells it as “100% stone ground from rich and hearty whole garbanzo beans,” so it brings protein and fiber along with the thickening, which is why it browns well. The Foods review also records the cost: pulses “present flavors described as ‘beany’ or ‘green’.” Some people taste it and some do not.

Blended silken tofu. Nisha Vora’s is the richest custard tested here: 12 ounces of silken tofu, 1 1/2 cups of unsweetened soy milk, 2 tablespoons of melted vegan butter, 5 tablespoons of brown sugar and 2 1/2 tablespoons of arrowroot, for ten big slices. Her reason for the tofu is specific: “Its natural proteins prevent the batter from spreading in the pan and help it cling to the bread.” Candice Hutchings blends medium or firm tofu instead of silken, which means a high speed blender is not optional on her version.

Chia. Dana Shultz uses one heaping tablespoon of chia seeds in 1 cup of almond milk and rests it in the fridge for 10 to 20 minutes before dipping. It is the only tested custard that needs a wait, and it is the only one built on the thinnest milk in this set.

Opaque beige custard falling from a blender jug into a wide shallow dish in a folded ribbon that holds a raised mound before it settles
The blended custard holds a ribbon. A thin whisked one lies flat and slides off the bread

Every route above is dairy free french toast by construction, and the milk you pick decides how well it browns. Silk publishes, per 1 cup, 8 grams of protein and 4.5 grams of fat for its original soymilk, against 1 gram of protein and 2.5 grams of fat for its original almondmilk. Eight times the protein, from the same brand, on the same serving size. That is the concrete reason four of the tested developers specify soy. Vora says it in one clause: “any sort of thinner, less fatty plant milk will make the batter thin.”

Lumps are the most common reported failure and they are a mixing order problem. On Andrews’s own recipe a reader named Vicki wrote that she “had a hard time getting the flour to mix into the Almond milk. It was very lumpy.” Andrews’s reply is the fix: add just enough milk to work the flour into a smooth paste first, then thin it with the rest. Another reader on the same page skips the whisk entirely and blends it.

Two ingredients get credited with more than they do. Black salt, kala namak, is a flavoring. The Foods review lists it among “flavoring agents” that “mimic the sulfur flavor of egg,” and nothing in this research says it helps the custard set, cling or brown. The published amounts are an eighth of a teaspoon for ten slices and a quarter teaspoon for eight, so start at the low end and taste. Nutritional yeast is the other one. Bragg describes it as having “a savory or umami flavor.” The eggy framing belongs to Andrews, who writes that it is in her recipe “for eggy flavor.” Take that as her palate rather than as a fact about the ingredient.

Flaxseed and baking powder both appear in Taylor’s recipe and neither has support behind it here. Bob’s Red Mill sells flaxseed meal on its omega-3 and fiber content and makes no thickening claim at all. Taylor says her teaspoon is there “to help thicken the batter,” which is her reason, not a mechanism. The half teaspoon of baking powder she includes comes with no explanation from her and none from anywhere else in this research. Leave both out or add them as her untested extras.

Bread, and how long to soak it

Two things about bread are settled and one is a real fight. Settled first: thickness. Vora tested 3/4 inch as the target with 5/8 inch still working, and she cuts bakery loaves at a full inch. Andrews explains the floor from the other end, that “soft, thinly sliced sandwich bread is an issue because as soon as you dunk it in the batter it will want to break apart.”

Also settled: what staling actually does. A 2021 paper in Current Research in Food Science ties crumb firmness directly to moisture loss during storage. Day old bread is not chemically different in any way that helps here. It is drier, and drier bread is firmer, and firm bread survives being soaked.

The fight is whether you want day old bread at all. Three tested developers say yes and give their loaves. Taylor uses one day old ciabatta and warns against soft sandwich bread. Turnbull calls for “sturdy, slightly stale bread.” Andrews uses sourdough that had been sitting a couple of days. Vora tested the opposite and got the opposite result: “you’ll get better results making French toast with fresh bread that’s been lightly toasted and dried out rather than day-old or stale bread.” Her method is an oven, 250 F for 6 minutes then 4 more after a flip for bakery bread, or 300 F for 8 minutes then 6 more for thick sandwich bread. They also disagree about sourdough specifically. Vora tells you to avoid it because it goes chewy and tastes too tangy. Andrews builds her recipe on it.

Both approaches are chasing the same thing, which is bread dry enough to take custard without collapsing. Pick the one that matches what is on your counter.

A square bread slice lying flat and fully submerged in pale beige speckled custard in a wide shallow bowl, seen from directly overhead

Soak time is where the tested numbers spread widest, so here they are side by side rather than averaged into one number nobody published.

Soak times as each developer published them, with the bread each one tested
Developer Bread Soak time
Nisha Vora Bakery bread or 3/4 inch slices 30 seconds a side
Nisha Vora Sandwich bread under 3/4 inch 20 seconds a side
Dana Shultz Sturdy bread 20 seconds a side
Dana Shultz Sandwich bread 25 to 30 seconds total
Nora Taylor Day old ciabatta 10 seconds a side
Alison Andrews Stale sourdough A few seconds a side
Sam Turnbull Sturdy, slightly stale Quick dip, no soak

Thirty seconds a side down to no soak at all, from seven tested instructions. The variable underneath is how dry and how thick the slice is, and Shultz is the only one who says so out loud: “If your bread is dry, leave it in a little longer.” Start at the short end for your bread and add time on the next slice.

The target has a name. Vora, citing America’s Test Kitchen, describes french toast that “should be soft but still have a thin layer of unsoaked bread in the middle (called the ‘bone’).” That thin dry seam is what keeps the slice from folding in the pan, and it is why soaking longer is not automatically better.

One technique from the wider cooking community is worth trying and is not tested for an egg free custard. On Seasoned Advice the highest scored answer on french toast quality soaks 30 seconds a side, then moves the slice to a rack for two more minutes so the custard spreads evenly through the crumb before cooking. No vegan developer in this set does that. Treat it as an experiment rather than an instruction.

Pan heat and getting real color

The tested heat settings do not agree either. Turnbull and Vora both say medium. Taylor and Hutchings say medium-high. Andrews says the pan should be “very hot” and gives no time at all. Shultz is the only one who names a number, a griddle at 350 F, which is also what Julie West gives as the standard griddle setting for french toast.

Medium is the setting to start on, and the reason is the browning mechanism above. Too hot and the surface passes color and goes to scorch before the inside of the slice has had the heat it needs to set. Too cool and the surface water never leaves, so the slice steams. Turnbull’s test needs no thermometer: flick a few drops of water at the pan, and “if the water sizzles and evaporates immediately, the pan is hot enough. If it just sits there and bubbles slowly, give it another minute or two.”

Per side, the tested times run 2 to 3 minutes (Turnbull, Taylor) or 3 to 4 minutes (Vora, Hutchings, West). Vora adds the adjustment that matters once you are past the first pan: drop the heat toward medium-low after the second batch, because the pan keeps climbing.

For the fat, she cooks with half vegan butter and half neutral oil, and gives her reason: “Even though vegan butter doesn’t contain milk solids, it still burned more easily during our tests.” Wipe the pan between batches so the browned residue does not carry over and go bitter.

Two slices in a black skillet, one an even mid bronze with a fine crackled surface and one still pale yellow with only a faint tan haze

The doneness test is the most useful single line in this entire research set, and it is Vora’s: “If you can push down on the center of the toast and see liquid squeezing out, it’s not done yet.” Nothing here gives an internal temperature for french toast and this page will not invent one. Press the middle and look.

On the chickpea flour science. A 2024 rheology study in Gels found chickpea flour sets in two stages, swelling between 40 and 70 C and then gelling near 95 C as its globulins denature. That is the reason a slice pulled off the pan early is thick but still wet in the middle: the starch has swollen and the protein has not set. The study measured flour dispersions in a rheometer, not bread in a skillet, so use it as an explanation and never as a target temperature.

Recipe: Vegan French Toast

Prep
5 minutes
Cook
15 minutes, 2 to 3 minutes a side
Total
20 minutes
Yield
4 to 6 slices

Ingredients

  • 1 cup plant milk, soy or oat
  • 2 tablespoons cornstarch
  • 2 tablespoons all purpose flour
  • 1 tablespoon maple syrup
  • 1 teaspoon vanilla
  • 1/2 teaspoon ground cinnamon, optional
  • Pinch of salt
  • 4 to 6 slices sturdy bread, 3/4 inch thick
  • Vegan butter and neutral oil for the pan

Method

  1. Put the cornstarch and flour in a wide shallow dish and add just enough of the milk to work them into a smooth paste, then whisk in the rest of the milk. Mixing in that order is what stops the lumps readers report.
  2. Whisk in the maple syrup, the vanilla, the cinnamon and the salt. The dish should be wide enough to lay a slice flat.
  3. Heat a skillet over medium with a piece of vegan butter and a splash of neutral oil. Flick a few drops of water at it: they should sizzle and go straight away.
  4. Dip a slice, both sides. Turnbull’s instruction for sturdy bread is a quick dip and no soak. For a thick bakery slice or a dry one, use the longer times in the soak table above.
  5. Cook two slices at a time, 2 to 3 minutes a side over medium, until the surface is an even bronze rather than shiny. Turn the heat toward medium-low once you are past the second batch.
  6. Press the center before you plate it. If liquid squeezes out, it goes back in the pan.
  7. Whisk the dish again before every slice. Cornstarch settles fast.
  8. Wipe the pan between batches and add fresh butter and oil. Hold finished slices on a wire rack in a 200 F oven so the undersides do not steam.

The quantities and the 5, 15 and 20 minute times are Sam Turnbull’s published ones for this recipe. The paste-first mixing order, the water test and the medium heat are hers too. The press test, the half butter and half oil, the heat drop after batch two and the 200 F holding oven are Nisha Vora’s. Everything else in the method is the range the seven tested recipes actually published, which is why the soak step points at a table instead of a number.

Swaps, and what each one costs

For a corn allergy, arrowroot goes in one for one. Bob’s Red Mill sells it as “a direct substitute for cornstarch for folks with corn allergies” with “no flavor of its own.” Vora uses it as her default thickener rather than as a substitution.

For gluten, Turnbull’s own swap is a gluten free all purpose blend plus gluten free bread, with her caveat attached: “since gluten-free breads vary widely, cooking time and texture may differ slightly.”

For a nut allergy, stay on soy or oat milk and skip the baked cashew version below. Chickpea flour is the other one to watch: on Andrews’s page a reader asked for an alternative because chickpea flour causes problems in her household, and the answer was a gluten free flour blend.

Without a pan, there are two tested routes. Turnbull air fries at 375 F for 8 to 10 minutes with a flip halfway, and says plainly that “they may not get as golden as pan-frying.” Richa Hingle bakes hers at 425 F for 15 to 18 minutes on a blended cashew custard, and pulls them once the edges of the custard and bread look golden. The baked one is a different dish presented under the same name, and it contains tree nuts.

For no oil at all, Turnbull’s route is a good nonstick pan with the butter and oil omitted.

The banana question is worth answering properly, because banana custard is the dominant published approach and it is the reason a lot of people go looking for something else. Turnbull is the strongest evidence here: she had a banana version on her site for years and replaced it, and her selling point for the new one is that it gives “that rich, custardy French toast texture without any banana flavor.” She kept the old one in her notes, one large banana with 3/4 cup of full fat coconut milk. There is a mechanical objection too, from Seasoned Advice: “bananas are too viscous to be absorbed by bread and won’t cook in a pan the same way that eggs do.” Banana coats rather than soaks. If you like banana on your french toast, Julie West’s version is one large banana to 3/4 cup of almond milk, cooked 3 to 4 minutes a side.

One honest limit, in Vora’s words: she has no tested substitute for silken tofu, suspects soft tofu would work, and has not tried a liquid vegan egg product to know the ratios. Nobody in this set has tested one either.

Fixing a soggy middle or a pale surface

A wet middle under a browned outside means the pan was too hot for the thickness of the slice, so the surface finished several minutes before the inside got the heat it needed. Drop the heat and give it longer, and judge it with the press test rather than the clock.

A wet middle with a pale outside means the slice took on too much custard for how dry the bread was. Cut the soak, and if the bread is soft, dry it in the oven the way Vora does before you dip anything.

A pale, faintly rubbery surface after a full four minutes usually means the custard has nothing in it to brown with. Thin almond milk with sugar and no protein source is the common version. Switch to soy, or add chickpea flour, or blend in tofu.

Custard sliding off in the pan is a thickness problem in the bowl. Whisk it again, and if it is still watery, work another spoon of starch into a little milk as a paste and stir that in.

Lumps that survive into the pan cook as gritty specks and there is nothing to be done about them once they are in the bread. Next batch, paste first.

An eggy smell that is too strong is black salt, and the only lever is how much went in. Published doses run from an eighth of a teaspoon for ten slices up to a quarter teaspoon for eight, so a batch that reeks of it is probably carrying two or three times what anyone published.

Storing and reheating cooked slices

The tested developers contradict each other here, so it is worth separating what they observed from what the food safety guidance says. Turnbull stores cooked slices in an airtight container for up to 3 days and freezes them in a single layer for up to 3 months. Taylor and Andrews both say their recipes do not store or freeze well and should be eaten fresh. Those are texture opinions, and none of them cites anything.

Plan on three to four days. Nebraska Extension’s window for refrigerated leftovers is “3 to 4 days,” with food back in the fridge within 2 hours and reheated “to a temperature of 165 F or until hot and steamy.” FDA’s chart sets the refrigerator at 40 F, and it is explicit that a freezer at 0 F “keeps food safe indefinitely,” so any 3 month freezer number is a quality claim rather than a safety one. Storage numbers on this site come from extension and FDA guidance rather than from recipe pages, which is written out in the editorial standards.

Reheating has two tested methods, both Vora’s. In a pan over medium with a little butter or oil, 1 to 2 minutes a side. In the oven, on a wire rack over a sheet pan, covered with foil, 8 to 10 minutes at 350 F, with the foil off for the last few if you want the outside crisp again. Turnbull reheats frozen slices straight from the freezer in a skillet, a toaster oven or a toaster.

Soaked but uncooked slices are the one thing nobody has tested. Two readers asked, one about the fridge overnight and one about freezing, and both developers said the same thing in different words: they had not tried it, and one called her answer a hunch. Andrews’s advice is to hold the custard instead of the bread and dip in the morning.

If you want to build a whole breakfast around this, vegan cinnamon rolls are the longer weekend version of the same craving, a vegan omelette runs on the same chickpea flour from the savory side, and the 7 day green smoothie challenge covers the mornings in between. There is a running list of every recipe on this site, and the comfort food recipes are where this one lives.

Sources