How to Roast Pumpkin Seeds in Oven: Convection, Rack Position and Doneness

By What Are Pepitas Editorial Team · Updated 2026-09-14 · Part of Pumpkin Seeds Recipe: Roasting In-Shell Seeds the Right Way

Roasting pumpkin seeds in the oven is mostly a moisture-management problem. An in-shell seed arrives with water trapped in the shell and in the kernel, and the job of the oven is to drive that water out before the shell browns too far. Every oven variable - convection or not, which rack, how thick the layer is, how full the oven is - changes the rate at which that moisture escapes, which is why the same recipe behaves differently from one kitchen to the next.

This page covers the oven side of the technique: what convection actually changes, where to put the rack, how thick the batch can be, and how to read doneness without guessing. For the cleaning and drying that come first, use how to roast pumpkin seeds; for the timing grid across temperatures, see how to bake pumpkin seeds. Everything is collected in pumpkin seed recipes.

Convection versus Conventional

A convection setting runs a fan that moves hot air across the food, which strips moisture from the seed surface faster and evens out the hot and cool zones in the oven. For in-shell pumpkin seeds, that is genuinely useful: it shortens the cook and gives a more uniform colour across the tray. The standard adjustment is to drop the temperature by 25 degrees F and check about 25 percent earlier than the conventional time, so a 325 degrees F conventional recipe becomes 300 degrees F convection, checked at minute 15 rather than 20.

There is one caveat. A strong convection fan can blow lightweight seeds around, especially if they are spread thin on parchment, which then curls and lifts. If your oven has an aggressive fan, use a rimmed bare sheet pan rather than parchment, and check once at the halfway point to make sure nothing has migrated to one corner. If your oven offers a low or bake-convection mode rather than full convection, use it and keep the original temperature.

Rack Position, Batch Thickness and Moisture Release

Use the centre rack for a single pan. It sits far enough from both the top element and the bottom element that the seeds dry rather than brown on one side. If you are running two pans, use the upper third and lower third positions and swap them, front to back and top to bottom, at the halfway point. Avoid the lowest rack position, where the pan base takes direct radiant heat from the element below and the undersides darken several minutes ahead of the tops.

Batch thickness matters more than total quantity. One loose layer with 5 mm of space between seeds is correct regardless of pan size; two overlapping layers will steam no matter how good your oven is. Moisture release is the mechanism: as water leaves the seeds it has to go somewhere, and in a crowded tray it condenses on the seeds above it. A slightly warm oven door cracked 2 cm for the first 10 minutes is an old trick that helps on very wet batches, though it costs you some temperature stability.

Oven load is the last variable worth naming. A single pan in an otherwise empty oven behaves differently from the same pan sharing space with two other dishes, because the extra mass absorbs heat and slows recovery every time you open the door to stir. If you are roasting seeds alongside something else, allow an extra 5 to 8 minutes and check the seeds on their own schedule rather than the other dish's. Never roast seeds on a rack below a pan of anything that might drip, since fat and moisture from above will ruin the batch.

Doneness Cues

Rely on three cues in combination rather than colour alone. First, sound: seeds begin to pop and crackle audibly as internal moisture flashes to steam, usually 3 to 5 minutes before they are done. Second, aroma: the smell shifts from green and raw to toasty and nutty, and that shift is your signal to start checking every 2 minutes. Third, colour: an even light golden brown with a few slightly darker patches, not a uniform deep brown.

Confirm with a cooled bite. Take three seeds from different parts of the tray - centre, edge, corner - since they cook at different rates, and let them sit for 5 minutes before biting. A finished seed cracks cleanly, tastes toasted rather than raw, and has no softness in the centre. If any of the three is soft, return the tray for 5 minutes at 300 degrees F and re-test.

Colour alone will mislead you in two specific situations. Seeds tossed in a dark spice such as smoked paprika or chile powder look finished about 5 minutes early, so rely on sound and aroma instead of sight. Seeds roasted on a dark-coated sheet pan also appear darker than they really are, because the pan colour changes how light reflects off the seeds; a light aluminium pan shows colour most accurately. When you are unsure, pull the tray and check with a cooled bite rather than trusting what you see through the oven door. Hulled kernels follow a much shorter curve entirely, so use how to roast pepitas in the oven if that is what you have.

• Centre rack for one pan; swap and rotate for two
• One loose layer only, 5 mm between seeds
• Convection: drop 25 degrees F, check 25 percent earlier
• Listen for popping 3 to 5 minutes before done
• Test seeds from centre, edge and corner of the tray
• Cool 5 minutes before the bite test

This page is part of the Pumpkin Seeds Recipe: Roasting In-Shell Seeds the Right Way guide.