A feeding trial was carried out to investigate the effect of various fish meals as dietary protein source on growth, feed utilization and body composition of Juvenile flounder. Ten experimental diets were prepared to contain different fish meals: 4 kinds of white fish meal (WM-1, 2, 3, 4), 3 kinds of herring meal (HM-1, 2, 3), mackerel meal (MM), WM mixture and HM+MM mixture. Dietary energy and protein levels were designed to be isocaloric ($3.8 kcal g^{-1}diet$) and isonitrogenous ($46\%$) by adjusting the levels of fish meal, wheat flour and squid liver oil. Three replicate groups of fish (initial mean weight: 11.1 g) were hand-fed to visual satiety two times daily for 7 weeks. Survival was not significantly different among all groups. Weight gain of fish fed the HM-1 and HM-3 diets was the highest, but not significantly different from that of fish fed the either WM-2 or HM+MM diets. The lowest weight gain was found in fish fed the W-1 and WM-3 diets. Feed efficiency and protein efficiency ratio of fish fed the WM-2, HM-1, HH-3, MM and HM+MM diets were significantly higher than those of fish fed other diets. Daily feed intake of fish fed the WM-3 diet was highest, but not significantly different from that oi fish fed the either WM-4 or WMM diet. Significant differences were found in contents of moisture, crude protein, crude lipid and ash of whole body of fish. The results of this study indicated that growth of flounder can be affected by dietary fish meal source and quality, and WM-2, HM-1, HM-3 and HM+MM mixture are considered as useful dietary fish meal sources under these experimental conditions.