“…Fruit weight had highly significant and positive association with fruit yield per vine (0.70), fruit length (0.64) and fruit width (0.27) and fruit length had significant and positive correlation with fruit yield per vine (0.48). Similar observations were made by earlier workers Yadagiri et al, (2017) for number of fruits per vine, fruit length, Rani et al, (2015) for number of fruits per vine, average fruit weight, pulp thickness in bitter gourd, Khan et al, (2015) for fruit length, average fruit weight, number of fruits per vine in bitter gourd and Yadav and Yadav (2015) for average fruit weight only at phenotypic level, in bitter gourd. 1.00 0.14 Critical r P value = 0.30 at 1% and 0.23 at 5 % *Significant at 5% probability level, **Significant at 1% probability level Where, X 1 =Vine length (m) X 2 =Number of branches per vine X 3 =Internodal length (cm) X 4 =Node at which first male flower appear X 5 =Node at which first female flower appear X 6 =Days to first male flower X 7 =Days to first female flower X 8 =Sex ratio X 9 =Number of fruits per vine X 10 =Fruit weight (g) X 11 =Fruit length (cm) X 12 =Fruit width (mm) X 13 =Fruit yield per vine (kg) 1.00 0.14 Critical r P value = 0.30 at 1% and 0.23 at 5 % *Significant at 5% probability level, **Significant at 1% probability level Where, X 1 =Vine length (m) X 2 =Number of branches per vine X 3 =Internodal length (cm) X 4 =Node at which first male flower appear X 5 =Node at which first female flower appear X 6 =Days to first male flower X 7 =Days to first female flower X 8 =Sex ratio X 9 =Number of fruits per vine X 10 =Fruit weight (g) X 11 =Fruit length (cm) X 12 =Fruit width (mm) X 13 =Fruit yield per vine (kg) Residual effect = 0.38 Bold diagonal value indicates direct effect Critical r P value = 0.302 at 1 per cent and 0.232 at 5 per cent probability level r p = Phenotypic path coefficients with fruit yield per vine (kg)…”