A feeding trial was conducted for 60 days to delineate the effect of both gelatinized and non‐gelatinized corn with or without supplementation of exogenous α‐amylase at two level of protein 35% (optimum) and 27% (sub‐optimum) on growth, tissue composition and protein sparing effect of carbohydrate in the diet of Labeo rohita fingerlings. Three hundred and sixty fingerlings (average weight 10.00±0.15 g) were randomly distributed in 12 treatment groups with two replicates for each. Twelve semi‐purified diets containing either 35% or 27% CP were prepared with different levels of α‐amylase (0, 50, 100 and 150 mg kg−1) and starch type [gelatinized (G)/non‐gelatinized, (NG)] viz. T1 (NG, 35% CP, 0 mg kg−1α‐amylase), T2 (G, 35% CP, 0 mg kg−1α‐amylase), T3 (NG, 27% CP, 50 mg kg−1α‐amylase), T4 (NG, 35% CP, 50 mg kg−1α‐amylase), T5 (G, 27% CP, 50 mg kg−1α‐amylase), T6 (G, 35% CP, 50 mg kg−1α‐amylase), T7 (NG, 27% CP, 100 mg kg−1α‐amylase), T8 (NG, 35% CP, 100 mg kg−1α‐amylase), T9 (G, 27% CP, 100 mg kg−1α‐amylase), T10 (G, 35% CP, 100 mg kg−1α‐amylase), T11 (NG, 27% CP, 150 mg kg−1α‐amylase) and T12 (NG, 35% CP, 150 mg kg−1α‐amylase). The fish were acclimatized to the experimental condition for 24 days during which control diet (NG, 35% CP) was fed. Weight gain percentage, specific growth rate, protein efficiency ratio and apparent net protein utilization of NG corn‐fed groups at 42.43% were significantly higher (P<0.05) than the same level of G corn‐fed groups. We conclude that 42.43% NG corn supplemented with 50 mg α‐amylase kg−1 feed at sub‐optimum level of CP (27%) significantly enhanced the growth and protein in L. rohita fingerling. Addition of dietary α‐amylase to G corn had no added advantage on growth of L. rohita fingerling.