Abstract: Most of feeding strategies which maximize the feeding rate are just designed for fish farming but not suitable for broodfish. Overfeeding caused by these feeding strategies will even impair the reproductive performance. However, moderate feed restriction can maintain fecundity while lowering breeding costs. Feeding strategies based on compensatory growth may promote growth and fecundity of broodfish, or make them mature faster. To test the effects on growth and reproductive capacity of female yellow catfish Pelteobagrus fulvidraco, fish were fed with four groups for 67 days: satiety feeding (AS), 80% satiety (80AS), 50% satiety (50AS) and starvation for 35 days and then satiety feeding for 31 days (HAS). Samples were taken at the middle (34 days) and the end of the experiment and various indexes were measured in this time. Results showed that the total food intake of the 80AS, 50AS and AS group roughly conforms to the proportion of the experimental design. The daily food intake of the HAS group was the highest. No significant difference was observed in weight gain rate and final weight between the 80AS group and the AS group. Meanwhile, the final weight of the HAS group was significantly lower than that of AS group, which shows the growth of the HAS group failed to reach those who fed at satiety level in the same period. As the serum lipid and protein concentrations increased in each group, the retention of lipid and protein in gonad increased. It can be inferred that parent fish preferentially transmit nutrition to gonads rather than muscles. There was no significant difference in final egg diameter between the AS group and 80AS group, which were significantly higher than the 50AS group and lower than the HAS group, indicating that the refeeding has a compensatory effect on egg growth. Feeding strategy did not impact absolute fecundity and fecundity, but alterd the expression of genes which related to gonad development. The relative expression of StAR, SF-1, 3β-hsd and cyp19a1 in the 80AS, 50AS and HAS was significantly lower than that in AS group in the middle stage. Nevertheless, no significant difference was found between these group and AS group in the end stage. The trend of plasma estradiol and testosterone was consistent with the above expression of genes, indicating that the synthesis and secretion of sex hormones and gonadal development in 80AS, 50AS and HAS groups were only inhibited by feeding strategy in the early stage. Finally, the inhibition effect was relieved, and the relevant indexes of each group recovered. These may be the result of interaction between the stress response and reproduction of broodfish. Based on the data above, satisfying the material and energy needs of gonadal development has priority in yellow catfish’s resource allocation. Although the growth of HAS group was lower than that of AS group, there was no difference in reproductive ability between HAS group and AS group, indicating that parent Pelteobagrus fulvidraco mainly compensated for fecundity rather than growth after starvation. Therefore, it is necessary to redesign the feeding strategy based on compensatory growth according to the physiological characteristics which need further exploration to improve the fecundity of Pelteobagrus fulvidraco broodfish. The growth and reproduction indexes of 80AS group were not significantly different from those of AS group, and the food intake was significantly lower than that of AS group, which was a more suitable feeding strategy in the breeding period of yellow catfish.