Abstract:The experiment was done at AB District Seed Farm, BCKV, Kalyani Simanta, West-Bengal, India during autumn-winter 2013-14 and 2014-15. The characters that exhibited higher Phenotypic and Genotypic Co-efficient of variation values were number of fruits per plant (76.86, 75.63%), fruit weight (43.88, 41.34%), harvest index (23.57, 22.29%), fruit yield per plant (53.61, 51.17%), anthocyanin in peel, total phenols and DPPH (2,2-diphenyl-l-picryl hydrazyl) free radical scavenging (FRS) capacity indicating that a greater amount of genetic variability was present for these characters which provide greater scope for selection. High heritability coupled with high genetic advance as percent of mean was observed for the characters like plant height, days to 1 st flowering, days to 50% flowering, number of fruits per plant, fruit weight, harvest index, fruit yield per plant, total sugar, anthocyanin in peel, total phenols and DPPH FRS capacity depicting that these traits were under the strong influence of additive gene action and hence simple selection based on phenotypic performance of these traits would be more effective. Fruit yield per plant showed highly positive significant correlation with number of primary branches per plant, number of fruits per plant, harvest index, vitamin-A and total phenols and significant negative correlation with days to 1st flowering, TSS, total sugars and total protein. Number of fruits per plant imparted the highest positive direct effect on yield followed by harvest index, fruit weight, days to 50% flowering and anthocyanin in peel. Number of fruits per plant and days to flowering were emerged as the main casual factors for positive or negative association of several characters with fruit yield per plant. Therefore, selection for fruit yield per plant based on these characters would be reliable.