- Survival of marine fishes during their early life history stage is tightly related to prey availability and predation pressure. Yet, our understanding of how individual larvae to entire assemblages are constrained by these factors is limited. We integrated biological sampling of larval fishes with fine-scale in situ imaging to relate patterns of larval fish feeding, growth, abundance, and assemblage structure to a gradient of fine- to regional-scale prey and predator conditions in the Straits of Florida. Otolith microstructure analysis showed that larval fish growth patterns reflected distributions of their preferred prey. Slower-growing reef fish (wrasse) larvae that are adapted to feed on patchy prey had faster growth at high prey densities, but also benefitted from dense patches of prey in otherwise low-density regions. In contrast, a fast-growing tuna species had faster larval growth at higher prey densities, but did not respond to prey patchiness, as their prey were consistently abundant. However, temperature appeared to strongly constrain larval tunas, evident from growth-temperature relationships that peaked at an optimum temperature, below their thermal limit. Further exploration of larval diet and growth of the abundant Thunnus atlanticus (blackfin tuna) revealed the combined consequences of poor prey availability and high temperatures. Larvae in high prey-low temperature conditions had fuller guts from eating higher quality prey and 30% faster growth rates, and were larger, older, and an order of magnitude more abundant, in comparison to fish in warmer temperatures. Predators had a strong effect on larval growth for both tunas and wrasses, evident by increasing strength of growth-selective mortality with predator densities or availability of dense patches of predators. Quantification of the entire plankton and larval fish community illustrated contrasting patterns of how larval fishes in coastal-upwelling compared to oligotrophic conditions respond to interannual prey variability. In the upwelling-influenced region larval fishes were tightly linked to fluctuations in prey, with overall 350% higher densities of larval fishes in high-prey versus low-prey conditions. This was due to numerous key larvae (e.g., tunas) having dramatic changes in abundance associated with fluctuations in their preferred prey. In contrast, larvae in the oligotrophic oceanic region were more consistent, as many larvae (e.g., wrasses) have flexible life history traits suggesting adaptive tolerance to a poor prey environment. Consistently strong predation pressure by gelatinous zooplankton may compound the effects of variable prey conditions in the continental region, as larvae in a poor-prey year experienced similar predation pressure as in high-prey conditions. Our results illustrate the importance of considering food web dynamics when predicting the response of larval fishes, and thus marine fish population replenishment, to ecosystem variability, particularly in light of ongoing global climate change.