The inherent complexity of yield and its environmental effects are the main constraints to yield improvement. An easier approach would be to improve the component traits that finally manifest into higher yield gains. Though a lot of literature has been generated in rice (Oryza sativa L.) on the type of gene actions, the success of breeding programmes in new plant types (NPTs) for yield has been hampered by lack of meaningful knowledge on the inheritance of yield components and yield contributing traits. Genes governing the character expressions in such altered genotypes were described. The F2 frequency distribution showed continuous variation indicating the polygenic nature of inheritance of these traits. In general, it was observed that dominant gene effects were of higher magnitude than additive gene effects in these crosses, though both were significant in the crosses studied. Thus, direct selection for yield per se will not be beneficial for genetic improvement and an indirect approach through selection indices will give better yield gains. Over-dominance and higher genetic heterosis for the characters, viz., plant height, filled grains per panicle, and grain yield per plant indicated that yield advantages could be easily exploited through heterosis. Also, multiple crossings in the F2 populations will allow more recombinations and prevent rapid fixation of alleles in small populations, and will help in producing wider variability, so that selection could be more effective in further generations.