Baked oatmeal squares live or die on two decisions: which oat, and what holds it together. Get those right and you can cut clean squares that travel in a lunchbox. Get them wrong and you’ve made a tray of warm porridge with ambitions.
I’ve made this in a lot of configurations over the past couple of years, mostly because my daughter kept asking for the muffin-tin version and I kept preferring the tray.
The Oat Question: Rolled, Quick or Steel-Cut
Rolled oats (old-fashioned, large flake) win, clearly. They’re steamed and pressed flat, so they drink liquid at a sensible pace and keep enough physical structure to hold a cut edge. A square made with rolled oats has visible flakes in it and a slight chew. That chew is the whole point.
Quick oats work, and they’re what you use if you want something closer to a soft cake. They’re rolled thinner and cut smaller, so they hydrate faster and the finished bake is denser and more uniform — less like oats, more like a bar. It slices very neatly. It also tastes a bit flat next to the rolled version, and I’d only reach for it if I were feeding someone who dislikes texture.
Steel-cut oats don’t work. Not “work less well” — don’t work. They’re chopped groats with the bran intact and no pre-steaming, and they need something like 25–30 minutes of simmering in liquid to soften. In a 35-minute bake at 180°C the surrounding custard sets long before the oat pieces have taken on water, and you get hard gritty bits suspended in something soft. I made this mistake once, served it to my husband without warning, and got a very diplomatic silence.
Binders Compared: Egg, Banana, Flax
The binder decides if you get squares or spoonfuls.
| Binder (per 200g oats) | Texture | Sweetness added | Slices cleanly? |
|---|---|---|---|
| 2 large eggs | Custardy, springy, holds a sharp corner | None | Yes — best of the three |
| 2 ripe bananas, mashed | Soft, moist, slightly gummy at the base | Roughly 25g of sugar’s worth | Warm: no. Fully cold: mostly |
| 2 tbsp ground flax + 6 tbsp water | Dense, a little tacky, nutty | None | Yes, but crumbles at the edges |
Egg wins if you want squares that survive a bag. Egg proteins coagulate into a continuous set network through the whole tray. Two eggs to 200g of oats is my standard.
Banana is the winner if you want it sweeter without added sugar, and it’s genuinely good — the pectin and starch do bind, just more softly. Two medium bananas let me drop the maple syrup from 80ml to 45ml. The catch is that a banana-bound bake must cool completely, at least 90 minutes, before you cut it. Cut it warm and you get lovely-tasting rubble. Same lesson as a banana bread loaf, really.
Flax is the vegan option and it’s fine. Ground flaxseed with water forms a mucilage gel that behaves a bit like egg white, but with no protein coagulation the set comes from thickening rather than structure. Squares hold if you’re gentle. Corners chip.
Liquid Choice: Dairy Milk, Plant Milk or Yogurt
Whole dairy milk is the default for a reason: fat and milk proteins both contribute to the set, so the bake firms up more than the same recipe made with thin plant milk.
Oat milk is the closest plant swap, and unsweetened almond is the furthest — almond milk is mostly water with a couple of percent solids, so a bake made with it needs an extra 10 minutes and still comes out looser. Soy milk sits between the two and actually sets well because of its protein content. If you’re going plant-based, use soy or oat, not almond.
Yogurt is the interesting one. Thinned with a splash of milk to a pourable consistency, it gives the tightest crumb of the three and a faint tang that makes the whole thing taste less like a dessert. Use 200g of plain yogurt plus 120ml milk in place of 300ml of milk. It browns slightly faster too, so check at 30 minutes.
Which do I actually use on a weekday? Milk, because it’s in the fridge. Yogurt when I’ve got Greek yogurt going spare, which is often, since I keep it for yogurt and granola parfaits.
One Dish or Muffin Tin
A 20cm square dish gives you nine squares with soft sides and a crisp top. A 12-hole muffin tin gives you twelve portions where every single one has a browned crust on five faces.
More crust sounds better. In practice the muffin version dries out faster — more surface area, more evaporation — and by day two they’re noticeably drier than tray squares from the same batch.
Muffin tin timing: 180°C fan, 22–25 minutes, and use paper cases or the oat will weld itself to the metal. Tray timing: 35–40 minutes.
I use the tray. The muffin tin is better only if portion control matters to you or the squares are going to a school where a tray gets mangled.
Ingredients
- 200g rolled oats (not quick, not steel-cut)
- 2 large eggs
- 300ml whole milk
- 80ml maple syrup or honey
- 50g butter, melted, or 4 tbsp neutral oil
- 1.5 tsp ground cinnamon
- 1 tsp baking powder
- 0.5 tsp fine salt
- 2 tsp vanilla extract
- 150g blueberries, chopped apple, or raisins
- 50g chopped walnuts or pecans, optional
Instructions
- Whisk eggs, milk, maple syrup, melted butter and vanilla in a large bowl until fully combined.
- Stir in the oats, cinnamon, baking powder and salt. Fold through the fruit and nuts.
- Leave the bowl on the counter for 20 minutes. The mixture will visibly thicken.
- Heat the oven to 180°C fan. Grease and line a 20cm square dish.
- Pour in, level the top, and press any exposed fruit down under the surface so it doesn’t scorch.
- Bake 35–40 minutes, until the top is golden and the centre barely wobbles when you nudge the dish.
- Cool in the tin for at least 30 minutes before cutting. An hour is better.
- Cut into 9 squares. Keeps 4 days covered in the fridge; reheat 40 seconds in the microwave.
Getting the Ratio Right
The number to hold onto is 1 part oats to 1.5 parts liquid by volume, counting eggs as liquid. In grams that’s my 200g oats to roughly 400ml of total wet ingredients (300ml milk, 80ml syrup, and two eggs at about 50ml each, minus a bit for the fat).
Go drier, say 1:1.2, and the bake is crumbly and tastes like undercooked granola. Go wetter, past 1:1.8, and it never fully sets — the excess water just sits there and steams.
Adjust for the add-ins. Fresh blueberries release a surprising amount of liquid; frozen ones release more, so pull the milk back to 270ml if you’re using frozen and adding them straight from the bag. Dried fruit does the opposite and steals water, so raisins want 20ml extra.
The 20-Minute Rest Before Baking
Twenty minutes on the counter, uncovered, before it meets the oven.
Dry oats hydrate slowly at room temperature, and if they go into a hot oven still dry, they carry on pulling water out of the custard while it’s trying to set. The result is a bake with two problems at once: gritty oats and a weepy base. Rest them and the oats arrive already softened, so the egg sets around a stable structure.
Twenty minutes is enough for rolled oats. If you rest overnight in the fridge, which works fine and is how I do it when I want breakfast ready fast, the texture goes softer and closer to overnight oats — pleasant, but less of a defined square. Add ten minutes to the bake time from cold.
Reading Doneness: The Wobble and the Thermometer
Press the centre with two fingers. It should feel set with the faintest give, and if you nudge the whole dish the middle should move as one piece rather than rippling. A ripple means liquid custard underneath.
The reliable check is temperature. An egg-set bake is done at about 88°C in the centre. Below 82°C it’ll sag when you cut it; past 93°C the egg starts squeezing out water and the base goes damp overnight in the fridge, which is the most annoying way to fail because it looked fine coming out.
Colour lies here. The top browns from the sugar in the maple syrup long before the middle is set, so a beautifully golden tray can be raw underneath — the same trap as pulling banana nut muffins too early on looks alone.
If you’ve overshot and the base is damp, put the cut squares back in at 160°C for 8 minutes on a wire rack. The exposed sides dry and you get most of it back.