Plasma levels of prolactin showed a pronounced annual cycle in free-living male great tits (Parus major). During the period from August to April, levels were very low. Prolactin levels started to increase in mid-April, and maximal levels were reached in June. By mid-July prolactin levels had decreased to near basal levels. The exact breeding stage was known for all males captured during the breeding period, and prolactin levels increased continuously from the period of territorial defense to the nestling period. Males were exposed to different light regimes at three different times of the year (late August, late November, and early March). Males exposed to 14L:10D (14 h light:10 h darkness) and 20L:4D showed pronounced prolactin cycles at all times of the year, but the patterns differed markedly with the season. In November the 20L:4D and the 14L:10D prolactin patterns differed markedly from each other. In the 20L:4D group prolactin levels started to increase before testes had reached maximal size, whereas in the 14L:10D group prolactin levels did not start to increase until testes were almost completely regressed. In early March the prolactin pattern of change over time was the same for great tits kept on 20L:4D and 14L:10D. In both cases prolactin levels increased during the testicular growth period, and prolactin levels were maximal during the period of spermatogenesis. Prolactin levels did not change over time in males kept on 8L:16D in August and November. Males exposed to short days in early March showed a significant increase in prolactin levels about 3 weeks after the onset of the experiment. Plasma levels of prolactin in males castrated in late November and exposed to a 20L:4D light regime did not differ from those in intact males. In castrated males given a testosterone implant prolactin levels immediately increased to significantly higher levels than those observed in intact or castrated males. Prolactin levels remained significantly higher in the testosterone implanted males for about a month. In one group of castrated birds the testosterone implant was removed 13 days after the onset of the experiment. This removal resulted in a significant decrease in circulating levels of prolactin.