In any crop improvement programme, for developing better genotypes the choice of suitable parents is a matter of great concern to the plant breeders. Therefore, it is imperative to know the role of important characters in the selection experiments aimed to increase seed yield. The present study has been conducted using Linear Discriminant function coefficient, Correlation between variables and discriminant scores and mean decrease in Gini index were used for computing relative importance of individual characters of Indian mustard discrimination between low and high oil content population of Indian mustard. Using the correlation between variables and discriminant score, the most important variables affecting the seed yield were secondary branches, primary branches and days to maturity. The three most important variables discriminating between oil content were siliqua length, secondary branches and seeds per siliqua.