A 7×7 complete diallel cross experiment was performed to study the genetic relationship of grain yield and some agronomic traits in rice. Analysis of variance showed significant differences among genotypes for all traits. Results of correlation analysis showed that rice grain yield was positively and significantly correlated with BY (rp = 0.81, rg =0.55), HI (rp =0.35, rg =0.79), NP (rp = 0.80, rg =0.67), PW (rp = 0.78, rg =0.83) and GSH (rp = 0.32, rg =0.81). Correlation between PH and GY (rp =-0.09, rg =-0.53) was negative at both type of correlation, but only rg was significant. The results of path analysis indicated that number of panicles (NP), panicle weights (PW) and plant height (PH) were strongly (92% of total variation) influenced grain yield (GY) as the predictor variables. Among these characters, PW reflected the highest direct effect (0.73) with grain yield; while minimum direct effect was indicated by PH (0.15). NP and PW can be used as effective selection criteria for grain yield increase. Factor analysis indicated that five important factors accounted for about 81.05% of the total variation among traits. The first factor assigned 23.0% of total variation between traits and was significantly related with grain yield, its components and flag leaf area. Other factors accounted for 18.65, 16.20, 14.20 and 9 percent of variation between traits. Applying genotype × trait (GT) biplot to the multiple trait illustrated that GT biplots graphically displayed the interrelationships among GY, PH, PW, HI, NP, GSH and BY. The results of this research indicated that use of NP and PW could be applied as a breeding strategy to obtain high yield rice cultivars.