In order to determine the influence of the main morphological traits on body weight of Macrobrachium nipponense, and to screen the optimal and measurable indices for artificial breeding of M. nipponense, the correlations between morphometric attributes and body weight of M. nipponense were explored using correlation analyses, path analyses and regression analyses. Ninety individuals from each of five populations (Tai Lake, Poyang Lake, Baiyang Lake, Weishan Lake and Dianshan Lake) were randomly sampled for the measurements of their body weight and 18 morphometric attributes, including the body length, overall length, rostrum length, second pereiopod length, diatance between eyes, carapace length, carapace width, carapace height, first abdominal segment height, first abdominal segment width, abdomen length, telson length, telson width, telson height, fantail length, fantail width, number of upper frontal eminencespin and number of lower frontal eminence spin. Correlation coefficients among all variables were calculated. Path coefficients were calculated by setting the body weight as a dependent variable and other attributes as independent variables via stepwise elimination. The path coefficients showed that: (1)five morphometric attributes of the Tai Lake population (body length, carapace width, second pereiopod length, carapace height, fantail length) significantly correlated to body weight. A multiple regression equation of the five attributes was established, where the regression intercept and regression coefficients were -9.878, 0.075, 0.314, 0.011, 0.320, 0.222, respectively. (2)seven morphometric attributes of the Poyang Lake population (carapace length, fantail length, telson width, carapace width, carapace height, telson height, body length) correlated significantly to body weight, where the regression intercept and regression coefficients were -1.618, 0.392, -0.280, 0.703, 0.524, -0.359, 0.688, -0.061, respectively. (3)four morphometric attributes of the Baiyang Lake population (body length, carapace width, second pereiopod length, fantail length) significantly correlated to body weight, where the regression intercept and coefficients were -4.796, 0.082, 0.222, 0.007,0.136, respectively. (4)four morphometric attributes of the Weishan Lake population (carapace length, first abdominal segment height, second pereiopod length, first abdominal segment width) significantly correlated to body weight, where the regression intercept and coefficients were -7.644, 0.248, 0.329, 0.025, 0.301. (5)five morphometric attributes of the Dianshan Lake population(body length, second pereiopod length, carapace width, telson width, frirst abdominal segment width) significantly correlated to body weight, where the regression intercept and coefficients were -6.257, 0.019, 0.018, 0.164, 0.264, 0.162, respectively. It is obviously that the body length and carapace traits in different populations have great effect on the body weight.High multiple-correlation coefficients of the reserved morphometric traits on the body weight in each population were greater than 0.85 (Tai Lake R²=0.965, Poyang Lake R²=0.904, Baiyang Lake R²=0.902, Weishan Lake R²=0.971, Dianshan Lake R²=0.955). The main attributes of each population effecting the body weight were confirmed. The clustering results showed that the closer the geographical location, the more similar the shape in terms of the morphological traits.