Summary. Angus and Angus crossbred heifers were ovariectomized, treated with oestradiol implants and randomly assigned to (1) the natural photoperiod of fall to spring for 43°N latitude or (2) extra light simulating the photoperiod of spring to fall. Weekly blood samples were taken for 6 months (fall to spring equinox). All heifers were cannulated every 4 weeks and blood samples were taken for 4 h at 15-min intervals. Sera were assayed for LH, FSH, prolactin and oestradiol. In samples taken weekly, serum LH and FSH concentrations were higher while serum prolactin was lower in heifers exposed to natural photoperiod. There was a photoperiod × time interaction for both FSH and prolactin with concentrations diverging as photoperiod diverged. Circulating concentrations of oestradiol were not different between groups. In samples taken every 4 weeks at 15-min intervals, baseline concentrations of LH and FSH and LH pulse amplitude were higher while prolactin pulse frequency was lower in heifers exposed to natural photoperiod. There was a photoperiod × time interaction for each of these pulsatile characteristics. The correlation between LH and prolactin concentrations estimated from the 15-min samples differed between the two photoperiod treatment groups. The pooled correlation coefficient (r) was −0·12 under natural photoperiod and +0·50 under extra light. There was also a photoperiod × time interaction with negative correlations occurring when photoperiod was decreasing and positive correlations occurring when photoperiod was increasing. These results support the hypothesis that photoperiod alters serum concentrations of LH, FSH and prolactin in cattle.