Generation mean analysis investigation in rice was carried out to assess the gene action operating in the inheritance of yield, its components, and quality traits. Five basic generations viz., P1, P2, F1, F2, and F3 of three crosses, namely IET-25475 x IET-25477(Cross-I), GNR-7 x GNR-3 (Cross-II), and GAR-13 x IET-25477 (Cross-III) were studied. A scaling test was applied to test the adequacy of the additive-dominance model for the traits under investigation. The result of individual scaling tests C and D revealed that the additive-dominance model was found inadequate for the description of variation in generation mean for all the eleven traits of all three crosses. Based on five parameter model, the main effects viz., mean (m), additive (d), and dominance (h) and two digenic interactions, additive x additive (i) and dominance x dominance (l) were significant for days to flowering in cross-I; for productive tillers per plant in cross-I and cross-III; for grains per panicle and 100-grain weight in cross-II; for grain yield per plant in cross-II and cross-III; for straw yield per plant and amylose content in cross-I and cross-II and days to maturity, plant height, L: B ratio and Zn content in all crosses indicated that involvement of additive, dominance as well as epistasis interaction for controlling this trait. BSA was carried out to identify markers linked to Zn content using 235 plants from F2 segregating generation of GAR-13/IET-25477, two indica ecotypes with extreme Zn content. The parents were screened for polymorphism using fifty rice microsatellite primer pairs. Out of 50 SSR primers, only two were found able to discriminate both the parents (HIGH Zn content and low Zn content) sufficiently. Two markers (RM 21 and RM 248) were identified as linked to the Zn content gene through BSA.