Influence of day length on seasonal endocrine responses were studied using stallions (seven per group). Treatments included 1) control, with natural day length; 2) 8 h light and 16 h dark (8:16) for 20 wk beginning July 16, 1982 then 16:8 from December 2, 1982 until March 5, 1984 (S-L); or 3) 8:16 from July 16, 1982 until March 5, 1984 (S-S). Blood was sampled hourly for 5 h every 4 wk; sera were pooled within horse, and luteinizing hormone (LH), follicle-stimulating hormone (FSH) and testosterone were quantified. Blood was collected every 20 min for 24 h every 8 wk and 2 wk before and after the December light shift. Samples were assayed for LH. Stallions in all groups underwent seasonal changes (P less than .05) in concentrations of LH, FSH, testosterone and basal concentrations of LH and amplitude of LH pulses. Season X treatment (P less than .05) reflected on early recrudescence of LH, FSH and testosterone concentrations in S-L stallions followed by earlier regression. Except for FSH hormone concentrations were depressed in S-S stallions. Number of LH pulses per 24 h was unaffected by season, treatment or their interaction. Mean amplitude of LH pulses was affected (P less than .05) by season X treatment; maximal values occurred in April vs February for control and S-L stallions, and minimal values occurred in December vs April. The season X treatment interaction (P less than .05) similarly affected basal concentrations of LH. Thus, seasonal changes in concentrations of LH, FSH and testosterone can be driven by photoperiod. Increased peripheral concentrations of LH during seasonal recrudescence of reproductive function apparently results from more LH secreted per discharge without an increased frequency of LH discharges.