Objective To determine the instar number and the basis for the division of developmental stages of Hyblaea puera larvae, thereby providing a theoretical foundation for the field control of this pest.

Method Based on the laboratory-reared specimens, the duration of each instar was recorded, and three morphological parameters: head capsule width, body length and body width, were measured. Frequency distribution histograms were constructed, and Crosby’s ratio, Brooks’ ratio, and regression analyses were applied to determine the optimal indicators for classifying the instar number and developmental stages of H. puera.

Result Frequency distribution analysis of head capsule width revealed five distinct distribution intervals, with mean values of 0.585, 0.996, 1.618, 2.375 and 3.173 mm, respectively. Both the coefficient of variation and Crosby's ratio for head capsule width across instars were less than 0.1. Linear, quadratic, and exponential regression models between head capsule width and instar number all showed good fits, with the quadratic model yielding the highest coefficient of determination (R²), indicating the best overall fit. Under conditions of a 14: 10 light-dark photoperiod and a temperature of 25 ± 1°C, the average durations of the 1st to 5th instar larvae were 1.433, 2.367, 2.533, 2.867, and 4.367 d, respectively.

Conclusion H. puera larvae undergo five distinct instars. Head capsule width conforms to both Dyar’s rule and Crosby’s rule and can therefore be considered a reliable indicator for instar determination.