A major goal in marine ecology is to understand patterns of larval dispersal and population connectivity. Dispersal plasticity allows for adaptive variation in dispersal phenotypes in response to variation in environmental conditions and may help to explain intraspecific variation in dispersal distances. However, this phenomenon has not yet been investigated in marine fishes. Here, we test the hypothesis that parents produce larvae with different dispersal-related traits in response to variation in environmental quality using the orange anemonefish, Amphiprion percula. By experimentally manipulating food rations, we show that parents produce larger offspring on low-food rations than on high-food rations. However, there was no effect of parental diet on larval critical swimming speed. We also found an effect on larval otolith core size which, in combination with parentage analyses, may provide a way to test the dispersal plasticity hypothesis in the field. This study shows that parents can produce different larval phenotypes in response to variation in environmental conditions, demonstrating plasticity in a dispersal-related trait which may help to explain observed variation in A. percula larval dispersal distances. Incorporating dispersal plasticity into our understanding of marine dispersal patterns may enhance our understanding of marine metapopulation ecology, fisheries management, and conservation.