Abstract: The experiment was conducted with 5 - 6 years old 52 seedlings from‘Katy’ ב Xinshiji’combination. The correlation and variation were studied among yield per plant, number of flowers per tree, fruiting rate of self-pollination, fertile flower rate, average fruit mass and the number of fruits per tree. The results showed that, ①there were wide variations in fruiting rate of self-pollination, fertile flower rate, average fruit mass and number of fruits per tree among apricot seedlings of F₁ populations, among them, variation coefficients of number of fruits per tree and fruiting rate of self-pollination were the biggest, 102.0%and 99.5% respectively, which showed that the selection potentiality of the two characters was large; ②number of fruits per tree, fruiting rate of self-pollination, fertile flower rate were most prominent positive correlation to yield per tree, correlation coefficients were 0.945, 0.541 and 0.439, respectively. Furthermore, fruiting rate of self-pollination and fertile flower rate were also most prominent positive correlation to number of fruits per tree. Then, it could be concluded that fruit setting percentage of self-pollination and fertile flower rate were the main factors to yield per plant; ③the correlation between average fruit mass and yield per tree was negative ( r =
- 0.176) , but itwas not significant, and its reason was that other factors effected reciprocally in the multifactor correlation analysis system. So, in this experiment the fruit size was not the main factor to yield per plant; ④according to the above results, we advanced primarily and discussed the selection method by two-steps of fertilitye-selection and quality-selection.

Key words: Apricot, F₁ population, Yield components, Variation, Correlations