A biochemically-based model was developed to simulate the growth, development, and metamorphosis of larvae of the Pacific oyster, Crassostrea gigas. The model defines larvae in terms of their protein, lipid, carbohydrate, and ash content and includes variation in growth efficiency and egg quality to better simulate cohort population dynamics. Changes in tissue composition occur as the larva grows and in response to the biochemical composition of the food. The premise behind this modeling study was that certain periods of larval life are more critical than others with respect to the availability of food and that food quality is as important as food quantity. The results of the simulations indicate that critical periods do exist, but that the period of larval life which is critical depends upon the composition of the available food supply and how it varies over time. Overall, the most critical time is late in larval life, near the time of metamorphosis. At this point, some variations in food quality are particularly efficacious, others particularly disastrous. But, under certain circumstances, events early or midway in larval life also dramatically change cohort survival. Simulations show that cohort survival varies in a relatively predictable way when salinity or food quantity vary. Both control time-integrated food supply to the larva by varying the amount of food ingested. Reduction of time-integrated ingestion reduces survival. Larvae with high growth efficiency are more successful, as are larvae coming from large eggs. The simple effect of timeintegrated food presents a stark contrast to the complexity introduced by varying food quality. Simulations indicate that it is late in larval life when larvae are most sensitive to changes in food quality. Increased protein at this time always improves survival. Increased lipid is most efficacious midway in larval life, but also exerts a positive impact late in larval life. Variations in carbohydrate are relatively inconsequential in affecting larval survival. Simulations in which food quantity and food quality vary independently show that cohort survival is sensitive to the exact timing and type of environmental change. Transient changes in food quantity influence survival primarily by varying the length of larval life. Transient changes in food quality, on the other hand, can produce large changes in survivorship by restricting the range of genotypes in the cohort that can survive, as well as by varying larval life span. The simulations support the adaptive advantage of larval cohorts with a relatively wide range of genotypes and suggest the important influence of variations in food quality in maintaining genetic variability.