Groups of photorefractory male European starlings have been kept under Nanda-Hamner LD schedules ranging from LD 6:6 to LD 6:54 for about 6 weeks. Subsequently, when transferred to LD 13:11, all birds of groups LD 6:18 and LD 6:42 went through a complete cycle of testicular growth and regression. Furthermore, they carried out a postnuptial molt, indicating that they have been able to terminate photorefractoriness under these LD schedules. In contrast, no positive responses could be detected in groups LD 6:6, LD 6:30, LD 6:36, and LD 6:54. Groups LD 6:12, LD 6:24, and LD 6:48 were intermediate in their reactions. Only part of these birds showed gonadal recrudescence when tested for photosensitivity. Changes in plasma testosterone levels were closely related to variations in testicular diameter. These results clearly suggest a strong circadian component in photoperiodic time measurement during the termination of photorefractoriness in the starling. Positive or negative responses were completely in line with predictions derived from the model of external coincidence. A comparison with the results of a similar experiment previously done with photosensitive birds failed to detect basic differences in the interpretation of the various LD cycles. These findings suggest that the same photoperiodic time measurement system is used at both phases of the starling's annual cycle.