The results of stages in controlled reproduction of African catfish C. gariepinus were investigated in females in six classes of body weight (I: >0 ≤1.5 kg; II: >1.5 kg ≤2.5 kg; III: >2.5 kg ≤3.5 kg; IV: >3.5 kg ≤4.5 kg; V: >4.5 kg ≤6.5 kg; VI: >6.5 kg), their ovulation being stimulated with Ovopel (1 pellet kg–1). The accepted classification significantly (P≤0.01) differentiated the weight of obtained eggs expressed both in grams and in percentage of female body weight. The highest weight of eggs was obtained from the heaviest fish (class VI) and the lowest from these of the lowest weight (class I) (1191.8 g and 218.7 g respectively) while in these two classes the obtained highest weight of eggs was expressed as the percentage of female body weight (14.21% and 20.19% respectively). The class of body weight significantly (P≤0.01) determined also the two investigated traits which described the quality of eggs. The lowest value of the least squares means for the percentage of fertilization and of live embryos was noted for the heaviest fish (class VI) (64.75% and 54.63% respectively) and the highest for the fish of the lowest weight, i.e., classes I and II (91.67% and 88.17%, 92.47% and 86.33%). In general the highest effectiveness of reproduction characterized the class of females whose body weight exceeded 4.5 kg but did not rise above 6.5 kg. In order to predict the weight of eggs [g], percentage of fertilization [12 h] and percentage of live embryos [24 h] within the investigated classes the regression equation was introduced using the weight of females as an independent variable. For each class regression equations were also derived in which the weight of females and weight of eggs were accepted as independent variables while the percentages of fertilization and of live embryos were used as dependent variables. The value of empirical data concerning the weight of eggs [g], fertilization percentage [12 h] and percentage of live embryos [24 h] were the basis for deriving a square function which permitted spatial prediction of the distribution of points of these variables on a plane – for each class of female body weight separately.