A two‐factor experiment was designed to determine the suitable dietary protein and lipid levels for juvenile Nibea diacanthus. Nine extruded pellet diets were formulated to contain three levels of protein (420, 470 and 520 g kg⁻¹) and three levels of lipid (70, 110 and 150 g kg⁻¹). Each diet was randomly fed to triplicate groups of 25 juvenile N. diacanthus (initial weight 12.12 ± 0.23 g) for 8 weeks in net cages. The results showed that weight gain rate (WGR), specific growth rate, final body weight and energy retention were significantly influenced by the dietary protein and lipid levels‐. The highest WGR (982.5g kg⁻¹.) of N. diacanthus was found in the group with dietary protein and lipid of 520 and 150 g kg⁻¹. Feed intake and feed efficiency were significantly impacted by the dietary protein levels. An interactive effect between dietary protein and lipid on the protein retention and protein efficiency ratio was observed. There were no significant differences in condition factor and survival among all treatments. Hepatosomatic index and viscerasomatic index of N. diacanthus were positively related with dietary lipid levels, but negatively with dietary protein levels. Crude protein, crude lipid, ash, moisture and energy contents of the whole body, muscle and liver were influenced by dietary protein and lipid levels. Moreover, total essential amino acid pattern of the muscle was correlated to those of dietary protein. Total protein concentration in the serum was affected by dietary protein and lipid levels. Meanwhile, both serum cholesterol and triglyceride concentrations increased with increasing dietary lipid levels. These results demonstrate that the diet containing 470 g kg⁻¹ protein and 110 g kg⁻¹ lipid is optimal for juvenile N. diacanthus and analysis of WGR by quadratic regression indicated that the estimated optimal protein‐to‐energy ratio for juvenile N. diacanthus was 24.53 mg protein kJ⁻¹.