Summary1. Mast seeding is a reproductive strategy in some perennial plants defined as synchronous production of large seed crops at irregular intervals. One widely accepted theory to explain this behaviour is the predator satiation hypothesis, which states that the synchronous and variable production of seeds within a population will maximize the probability of seed survival through satiation of seed predators. 2. Although some short-term studies have documented the influence of variable and synchronized production of seeds on herbivore attack rate during one or few mast years, long-term data including multiple mast seeding years and patterns of cone production and herbivore attack on individual trees are needed to assess (i) how cone production, variability and synchrony affect individual plant fitness and (ii) the functional responses of seed predators to mast seeding events. 3. We tested these objectives, collecting long-term (29 years) data on female seed cone production and rates of seed predator attack from 217 individual contiguous trees within a Pinus ponderosa population. 4. Our results support the predator satiation hypothesis. First, we found high interannual synchrony and variability in seed cone production and a type II functional response of seed predators to available cones. Secondly, years with high cone production (mast years) had markedly lower rates of seed predator attack than years of low production (i.e. a population-level satiation effect). Thirdly, within mast years, individuals with high cone production had markedly lower rates of attack than individuals with low cone production (i.e. an individual-level satiation effect). Finally, individual trees with greater synchrony and more variable cone production suffered lower rates of attack. 5. Synthesis: Our long-term data on individual trees lend strong support to the hypothesis that mast seeding in Pinus ponderosa has evolved in response to natural selection from insect seed predators.