The intensity of larval diapause in Sesamia nonagrioides Lef (Lepidoptera: Noctuidae) was investigated under laboratory conditions. Newly hatched larvae were exposed to different stationary photoperiods (from LD 7:17h to LD 14:10 h), at a constant temperature of 25°C. Diapause incidence was higher when larvae were exposed to daylengths shorter than the critical value (LD 12:12 h), whereas the within-treatment variation in the larval period appeared to be significantly correlated with the photoperiod applied. The incidences of diapause and the duration of larval development were also measured after exposing larvae to short photoperiods (LD 8:16 h, LD 10:14 h or LD 12:12 h) in combination with various temperatures (20, 22.5 or 25°C). Although an increase in the incidence of diapause appeared with the lowering of the temperature, no statistical differences were observed in the time needed for pupation within the photoperiodic treatments at the temperatures of 20 and 22.5°C. Furthermore, when diapausing larvae were transferred to the long photoperiod of LD 16:8 h, they immediately proceeded to pupation, regardless of the photoperiod or the temperature to which they had been previously exposed, indicating that there were no differences in the intensity of diapause. Photoperiodic changes from LD 10:14 h to LD 12:12 h or to LD 14:10 h at different larval ages reduced the intensity of diapause with (a) early age of transfer and (b) increase of daylength. By contrast, when larvae were transferred from the long photoperiod of LD 14:10 h to shorter, such as LD 10:14 h or LD 12:12 h, a small increase in the intensity of diapause with the shortening of the daylength was apparent. These results support the hypothesis that insects may compare the duration of the photoperiod and could classify them as either longer or shorter in relation to the critical value.