A $3{ imes}2$ factorial experiment was conducted to investigate the proper dietary protein and lipid levels for the growth of sub-adult flounder Paralichthys olivaceus reared during the summer season. Six experimental diets were formulated to contain three levels of protein (45%, 50% and 55%) and two levels of lipid (9% and 14%). Duplicate groups of fish (initial body weight of 298 g) were hand-fed to apparent satiation during the summer season ($21.8{pm}1.7^{circ}C$) for 60 days. Survival of each group was over 83% and there was not significant difference among all groups. Weight gain of fish fed the 45% protein diet with 14% lipid was not significantly different from that of fish fed the 50% and 55% protein diets with 9% and 14% lipids, but weight gain of fish fed the 45% protein diet with 9% lipid was significantly lower than that of fish fed the 55% protein diets with 14% lipid. Feed efficiency tended to increase with increasing dietary lipid level at each protein levels, although no significant differences were observed at 50% and 55% protein levels. Protein efficiency ratio, daily feed intake, condition factor and hepatosomatic index were not significantly affected by dietary protein and lipid levels. Crude lipid content of the liver tended to increase with increasing dietary lipid level at the same protein levels, but the opposite appearance was found for moisture content. The contents of moisture, crude protein and crude lipid of the dorsal muscle were not significantly affected by dietary protein and lipid levels. Based on data obtained form this study, inclusion of dietary protein at level of 45% appears sufficient to support optimal growth, and an increase of dietary lipid level from 9% to 14% has beneficial effects on feed utilization of sub-adult flounder during the summer season.