We have previously shown that reproductive seasonality of bucks was prevented for 2 consecutive years by short photoperiodic cycles. To determine the effect of the length of treatment time on bucks subjected to the same photoperiod conditions, experiments were continued for a third consecutive year on 3 groups of 6 Alpine and Saanen bucks. The control group was kept under natural photoperiodic conditions, while the experimental groups were exposed alternately to 1 month of long days and 1 month of short days (group 2M) or to 2 months of long days and 2 months of short days (group, 4M). Prolactin profiles indicated that bucks from both experimental groups responded adequately to rapid photoperiod changes as their plasma prolactin levels were significantly higher in long days (mean +/- SEM; 2M: 61.1 +/- 15.9 ng/ml; 4M: 102.2 +/- 13.5 ng/ml) than in short days (2M: 35.3 +/- 8.2 ng/ml; 4M: 46.1 +/- 9.0 ng/ml). Testosterone secretion was also dependent on day length (P < 0.0001), since testosterone concentrations of experimental animals were higher during long days (2M: 7.0 +/- 0.7 ng/ml; 4M: 10.2 +/- 1.1 ng/ml) than during short days (2M: 4.3 +/- 0.4 ng/ml; 4M: 5.0 +/- 0.9 ng/ml). Furthermore, controls displayed a high level of sexual behavior (always higher than 10%) and the proportion of bucks unable to ejaculate was significantly lower (P < 0.01) than the experimental animals (2M: 25.6%; 4M: 28.1%).(ABSTRACT TRUNCATED AT 250 WORDS)