In order to study correlations between traits and its analysis to cause and effect relationships in rapeseed plant, an experiment was conducted by a randomized complete block design with three replications for 8 cultivares of fall rapeseed in 2006-2007 growing season at the research field of Abooreihan Campus, University of Tehran, Pakdasht. Evaluation and analyze for 18 important traits in statistical analysis. Results showed significant differences among cultivars for all traits except stem weight, pod length, LAI in canopy closure, height of the first branch, number of days to the beginning of branching. Positive significant correlation was observed between seed yield and biological and oil yields, stem weight, harvest index, oil percent, plant height and number of pods/plant. But, negative correlation was observed between seed yield with number of days to the beginning of branching, number of days to the beginning of flowering, number of days to 50% flowering, number of days to full flowering. According to stepwise regression, seed yield trait is considered as dependent variable and biological yield, stem weight, number of pods/plant, LAI in canopy closure, plant height and number of days to 50% flowering traits as independent variables. Model determination coefficient is R2=1. Negative direct effect of weight yield and plant height on seed yield were canceled by positive indirect effects via biological yield and number of pods/plant and caused that correlation increase between this traits and grain yield. Conversely, positive direct effects of number of days to 50% flowering on seed yield were canceled by negative indirect effects of the same two traits. Therefore, the most important traits orders to selection index for seed yield improvement were biological yield and number of pods/plant, respectively.