Mast seeding involves the episodic and synchronous production of large seed crops by perennial plants. The predator satiation hypothesis proposes that mast seeding maximizes seed escape because seed predators consume a decreasing proportion of available seeds with increasing seed production. However, the seed escape benefits of masting depend not only on whether predators are satiated at high levels of seed production, but also on the shape of their functional response (type II vs. type III), and the actual proportion of available seeds that they consume at different levels of seed production. North American red squirrels (Tamiasciurus hudsonicus) are the primary vertebrate predator of white spruce (Picea glauca) mast seed crops in many boreal regions because they hoard unopened cones in underground locations, preempting the normal sequence of cone opening, seed dispersal, and seed germination. We document the functional response of cone-hoarding by red squirrels across three non-mast years and one mast year by estimating the number of cones present in the territories of individual red squirrels and the proportion of these cones that they hoarded each autumn. Even though red squirrels are not constrained by the ingestive and on-body (fat reserves) energy reserve limitations experienced by animals that consume seeds directly, most squirrels hoarded <10% of the cones present on their territories under mast conditions. Cone availability during non-mast years also reached levels that satiated the hoarding activity of red squirrels; however, this occurred only on the highest-quality territories. Squirrels switched to mushroom-hoarding when cone production was low and mushrooms were abundant. This resulted in type III functional response whereby the proportional harvest of cones was highest at levels of cone availability that were intermediate within non-mast years. Overall, more cones escaped squirrel cone-hoarding during a mast event than when cone production was low in non-mast years, which supports the predator satiation hypothesis. However, the highly variable seed escape in non-mast years may help to explain why all spruce cone production is not concentrated into fewer, larger, mast years.