A 60 days feeding experiment was carried out with Black Sea turbot Psetta maeotica to determine the amount of poultry by-product meal (PBM) that could replace fish meal (FM) in formulated diets without reducing growth performance. Juvenile Black Sea turbot (initial average weight, 30 g) were fed five isoenergetic (gross energy, 20.5 +/- 0.21 kJ g(-1) diet) and isonitrogenous diets (protein content, 550 +/- 0.35 g kg(-1)). The control diet used white FM as the sole protein source, the other four diets were prepared to replace FM protein at levels of 25%, 50%, 75% and 100% with PBM. The fish readily accepted all experimental diets and no mortality were recorded during the trial. There were no significant differences in growth performance of turbot (P < 0.05) fed the diets with 25% and 50% replacement levels compared with fish offered the control diet (100% FM), however, final body weight and specific growth rate values in the 50% replacement diet were about 8% lower than those of the control. Total nitrogen excretion in fish fed 50% replacement diet were about 10% higher than the control group, even though these parameters were not found to be statistically different. At the levels of 750 and 1000 g kg(-1) of the protein, PBM inclusion caused a severe decrease in growth performance, feed utilization, protein efficiency ratio and per cent nitrogen retention. The results in the present study indicate that up to 25% of FM protein can be replaced by PBM protein without causing reduction in growth performance, nutrient utilization and nitrogen retention.