This experiment was conducted to determine the each level of a newly developed squid liver powder (SLP) as a dietary protein source replacing fish meal (FM) in juvenile flounder. A diet excluding SLP (FM diet) and a commercial diet were two control diets. Six experimental diets were formulated to replace FM with three types of SLP at 5 and 10% level, respectively (SLP-A5, SLP-A10, SLP-B5, SLP-B10, SLP-C5, SLP-C10). Prior to the start of a feeding trial, fish were fed the commercial diet for 1 week to adjust to the experimental conditions. Fish with average weight of $22.8{pm}0.40g$ ($mean{pm}SD$), were randomly distributed to each aquarium as groups of 20 fish and fed the experimental diets in triplicate at satiation for 8 weeks. After 8 weeks, weight gain and specific growth rate of fish fed SLP-C10 was lower than those of fish fed FM diet and SLP-B5. Meanwhile, no significant difference was observed in WG and SGR among fish fed diets excluding SLP-C10. Feed efficiency (FE) and protein efficiency ratio (PER) of fish fed each level of three types of SLP were not different from those of fish fed FM diet. However, fish fed SLP-C5 and SLP-C10 has lower FE and PER than those of fish fed a commercial diet, SLP-A5 and SLP-B5. Therefore, these results indicate clearly that dietary SLP-B could be a good protein source to develop inexpensive commercial diet in juvenile flounder.