The breeding season of wild starlings is controlled by photoperiod. Full breeding condition is attained during exposure to lengthening days in the spring, and photorefractoriness ensues. The reproductive system of starlings will not subsequently be stimulated by long day lengths until photorefractoriness is dissipated by the short day lengths experienced in the autumn and winter. Unlike most studies on avian photoperiodism, this investigation involved manipulation of light intensity of a fixed photoperiod rather than of photoperiod itself. Photosensitive starlings transferred from short days to long days of different light intensities underwent graded reproductive responses according to the light intensity they experienced. Testes size in the group in the lowest intensity (3 lux) increased faster than that in controls on short days of normal intensity, but they did not become photorefractory. Testes size increased in the groups on 13, 45, and 108 lux and subsequently became photorefractory. However, the 13- and 45-lux groups required more time to become photorefractory than did the 108-lux group. The responses observed were similar to those seen in starlings exposed to different photoperiods (e.g., 11 h light:13 h dark [11L:13D], 13L:11D, 16L:8D, 18L:6D), even though all were on the same 18L:6D photoperiod. Initially, the results appear to challenge the external coincidence model for photoperiodic time measurement, but consideration of the phase response curve of the circadian rhythm of photoinducibility in starlings and the way in which it might be affected by low light intensities refute this challenge.