Tarte aux amandes was always mostly almonds. Take the pastry flour out entirely and you get a dessert that sets like marzipan, carries ten grams of protein a slice, and tastes like something you ordered rather than something you were allowed.

Watch: the flourless French almond tart, start to finish.

There is a particular kind of disappointment that comes from a flourless dessert. It arrives looking apologetic. It is dense in the wrong way, or sweet in the wrong way, or it has been sweetened with something that tastes faintly of the health-food aisle.

This is not that.

The reason is structural. A classic French tarte aux amandes is already built around ground almonds — the frangipane that fills it is almonds, butter, sugar and egg, with a spoonful of flour added mostly out of habit. The pastry shell underneath is where the wheat actually lives. Remove the shell, rebuild it out of almonds, drop the spoonful of flour from the filling, and nothing is missing. The tart does not need rescuing with starch. It never needed the starch in the first place.

What you get instead is a tart with two textures that belong together: a thin, sandy, press-in crust that behaves like shortbread, and a centre that sets somewhere between custard and marzipan — close-grained, faintly chewy at the edges, still yielding in the middle. The flaked almonds on top go deep bronze and shatter under the knife.

It takes one bowl more than it strictly needs, no rolling pin, and no chilled-dough ritual. You press the crust in with your knuckles.

What makes it worth knowing

Three things, and only the third is about nutrition.

It is grain-free by construction, not by substitution. No wheat flour, no rice flour, no cornstarch, no gum. Nothing is standing in for anything. That matters for texture — most flourless baking fails because a substitute is doing a job it is bad at — and it matters if you cannot eat gluten, because there is nothing to omit.

It holds. The tart is better on day two than on the hour it comes out of the oven, which makes it a genuinely useful thing to make for people. Bake it the night before. It is one of the few desserts that improves while you sleep.

And the macros are unusual for a dessert. About 10 g of protein a slice, 5 g of fibre, and roughly 21 g of carbohydrate — a fraction of what a conventional pâte sucrée tart of the same size carries, because the single largest carbohydrate source in that tart is the shell you just replaced. A slice also brings close to 10 mg of vitamin E, most of a day's requirement, and about 110 mg of magnesium. Almonds are one of the richest whole-food sources of both.

None of that makes this health food. It is 430 calories of butter and almonds and it is supposed to be.

The evidence, honestly

Viva sorts health claims into three tiers. Here is where this one sits.

ESTABLISHED. Eating almonds regularly lowers LDL cholesterol. A systematic review and meta-analysis of 15 randomized controlled trials found almond consumption reduced LDL cholesterol by about 5.8 mg/dL and total cholesterol by about 10.7 mg/dL against control diets. A separate meta-analysis in people with type 2 diabetes found a comparable LDL reduction of about 5.3 mg/dL. These are modest, real, repeatable effects from a daily handful — not from a slice of tart.

ESTABLISHED, and it cuts against us. You may have read that almonds contain fewer calories than the label claims. That is true, and it has been measured carefully: in a controlled crossover trial, whole almonds delivered 4.6 kcal/g rather than the 6.0–6.1 kcal/g the standard Atwater calculation predicts — a 32 per cent overestimate. The almond's cell walls trap some of the fat and it leaves the body unabsorbed.

Here is the part that usually gets left out. The same research group tested almonds in different physical forms and found the effect depends entirely on how broken-down the almond is. Whole natural almonds measured 4.42 kcal/g. Chopped almonds, 5.04. Almond butter measured 6.53 kcal/g — statistically indistinguishable from the predicted value. Grinding ruptures the cell walls and hands the fat over.

Almond flour is ground almonds. So in this tart, you are absorbing essentially everything on the label. The nutrition panel below is not secretly generous. We would rather tell you that than let a flattering statistic do work it cannot do.

EXPERIENTIAL. That it keeps people satisfied with a smaller slice than a flour tart would. That is our kitchen's observation across a lot of testing, not a finding.

At a glance

Serves 10
Active time 25 minutes
Chilling 20 minutes
Baking 50 minutes total
Cooling 1 hour, minimum
Tin 23 cm (9 in) loose-bottomed tart tin
Make ahead Yes — better on day two, keeps 4 days
Grain-free Yes. No wheat, no starch, no gum

Ingredients

For the almond crust

  • 200 g almond flour (finely ground blanched almonds)
  • 45 g cold unsalted butter, cubed
  • 30 g maple sugar, or coconut sugar
  • 1 large egg yolk
  • 1 tsp vanilla extract
  • ¼ tsp fine sea salt

For the frangipane

  • 170 g almond flour
  • 100 g unsalted butter, well softened
  • 80 g maple sugar, or caster sugar
  • 2 large eggs, at room temperature
  • 1 large egg yolk
  • 1 tbsp dark rum
  • ¾ tsp pure almond extract
  • Finely grated zest of 1 orange
  • ¼ tsp fine sea salt

To finish

  • 40 g flaked almonds
  • 2 tbsp apricot preserve, warmed with 1 tsp water

A note on the almond flour. Use blanched, finely ground almond flour, not almond meal with the skins left in. Meal makes a coarser, darker, slightly bitter tart. Both work; they are not the same dessert.

A note on the rum. It is not for the alcohol. Almond and rum share aromatic compounds, and a tablespoon makes the almond read as deeper rather than sweeter. Brandy or amaretto do the same job. If you leave it out, add an extra ¼ tsp of almond extract and the zest of half a lemon.

Method

  1. Make the crust. Heat the oven to 175°C (350°F). In a bowl, rub the cold butter into the almond flour, maple sugar and salt with your fingertips until the mixture looks like damp sand with a few pea-sized pieces left. Add the egg yolk and vanilla and work it briefly with a fork, then with your hand, until it comes together into a soft, slightly greasy dough. It will not behave like pastry. It should not.
  2. Press and chill. Tip the dough into the tart tin. Press it out from the centre with your knuckles, working it up the sides to an even 4–5 mm, and square off the top edge with your thumb. Prick the base in a dozen places with a fork. Chill, uncovered, for 20 minutes — this is the step that keeps the sides from slumping.
  3. Blind bake. Bake the shell for 12–14 minutes, until the edges are set and just beginning to colour. It will look pale and underdone in the centre; that is correct, it goes back in. Leave it in the tin and set it aside. Do not use baking beans — there is no gluten to shrink and the weight only compresses the crust.
  4. Make the frangipane. Beat the softened butter with the maple sugar and salt for 2–3 minutes, until pale and aerated. Add the eggs and yolk one at a time, beating well between each. It may look split or curdled after the second egg. Ignore it. Add the almond flour, rum, almond extract and orange zest, and fold until just smooth and glossy — 20 seconds, no more. Overbeating now gives you a tart that rises and cracks.
  5. Fill and top. Scrape the frangipane into the shell and level it with an offset spatula, taking it right to the crust. Scatter the flaked almonds over the surface in an even layer and press them down very lightly so they anchor.
  6. Bake. Bake for 35–40 minutes. It is done when the top is deep golden, the almonds are toasted, and the centre has a faint wobble when you nudge the tin — like a set custard, not a liquid one. A skewer in the middle should come out with a few moist crumbs clinging to it. If the almonds darken before the centre sets, lay a sheet of foil loosely over the top for the last 10 minutes.
  7. Glaze and cool. Warm the apricot preserve with the water until loose, push it through a small sieve, and brush it thinly over the hot tart. Cool in the tin for at least an hour before unmoulding — the centre firms as it cools and a warm tart will not slice cleanly.

Serve at room temperature, in thin wedges, with crème fraîche and nothing else.

Where it goes wrong

The crust slumped down the sides. It went into the oven warm. Twenty minutes in the fridge is not optional, and in a warm kitchen give it thirty.

The middle sank. Overbeaten frangipane. Once the almond flour goes in, fold — do not whip air into it.

It cracked across the top. Oven too hot, or baked too long. Check it at 35 minutes.

It tastes faintly of marzipan and nothing else. Not enough salt or zest. Both quarter-teaspoons of salt matter, and the orange is doing more than you think.

It was greasy underneath. The butter in the crust was too soft when you rubbed it in. Start with it cold and cubed, straight from the fridge.

Substitutions

Instead of Use What changes
Maple sugar Caster sugar, coconut sugar Caster is cleaner and sweeter; coconut adds a faint caramel note
Dark rum Amaretto, brandy, or omit See the note above
Flaked almonds Flaked hazelnuts, pistachios Pistachio is beautiful and costs more
Apricot glaze Warmed honey, or nothing The glaze is mostly for shine
Orange zest Lemon, or 1 tsp orange blossom water Lemon is sharper and more austere

To make it dairy-free: replace all the butter with a good cultured vegan block, not a spread. The crust will be slightly more fragile.

To push the protein higher: fold 15 g of unflavoured whey or collagen into the frangipane with the almond flour. It adds about 1.5 g of protein a slice and makes the centre firmer and less tender. We prefer it without.

Nutrition

Per slice, based on 10 slices.

Energy 430 kcal
Protein 10 g
Carbohydrate 21 g
of which fibre 5 g
of which sugars 12 g
Fat 34 g
of which saturated 9 g
Sodium 135 mg
Vitamin E approx. 10 mg
Magnesium approx. 110 mg

Estimated from ingredient weights and standard composition data. As discussed above, this is not a case where the almond calories quietly fail to count — the almonds here are ground.

Questions, answered

Can you make frangipane without flour? Yes. The flour in a classic crème d'amandes is a small amount added for insurance, not structure — the egg and the ground almonds set the filling on their own. Leaving it out gives a slightly denser, more marzipan-like centre.

What can I use instead of pastry for a tart base? A press-in almond crust: almond flour, cold butter, a little sugar, one egg yolk and salt, pressed into the tin with your hands and blind-baked for 12–14 minutes. It needs no rolling, no resting and no baking beans.

Is almond flour lower in calories than wheat flour? No. Almond flour is higher in calories and fat per 100 g than wheat flour, and lower in carbohydrate. What changes in a tart like this is the carbohydrate load and the protein and fibre that come with it, not the calorie count.

Do almonds really have fewer calories than the label says? Whole almonds, yes — measured at about 4.6 kcal/g against a predicted 6.0–6.1. But the effect depends on grinding: almond butter measured 6.53 kcal/g, matching the prediction. Almond flour behaves like the ground form, so the discount does not apply here.

Why did my almond flour crust slump? It was not cold enough going into the oven. Chill the pressed shell, uncovered, for a full 20 minutes — 30 in a warm kitchen.

How long does an almond tart keep? Four days, covered, at room temperature. It is better on day two than on the day it is baked.

Sources

Verified via PubMed.

  1. Lee-Bravatti MA, Wang J, Avendano EE, King L, Johnson EJ, Raman G. “Almond Consumption and Risk Factors for Cardiovascular Disease: A Systematic Review and Meta-analysis of Randomized Controlled Trials.” Advances in Nutrition, 2019;10(6):1076-1088. DOI
  2. Novotny JA, Gebauer SK, Baer DJ. “Discrepancy between the Atwater factor predicted and empirically measured energy values of almonds in human diets.” American Journal of Clinical Nutrition, 2012;96(2):296-301. DOI
  3. Gebauer SK, Novotny JA, Bornhorst GM, Baer DJ. “Food processing and structure impact the metabolizable energy of almonds.” Food & Function, 2016;7(10):4231-4238. DOI
  4. Moosavian SP, Rahimlou M, Rezaei Kelishadi M, Moradi S, Jalili C. “Effects of almond on cardiometabolic outcomes in patients with type 2 diabetes: A systematic review and meta-analysis of randomized controlled trials.” Phytotherapy Research, 2022;36(5):1839-1853. DOI

Nutrient composition estimated from ingredient weights against USDA FoodData Central values for raw almonds, butter and eggs.