The purpose of these experiments was to determine whether daily gonadotropin surges that occur in intact or ovariectomized hamsters kept in short days (less than 12 h of light/day) are manifest because of extremely low levels of steroids. Hamsters were ovariectomized and placed in 16L:8D or 10L:14D. After 10-13 weeks, animals in each photoperiod were divided among four treatment groups: estradiol implant plus progesterone injection 5 days later; estradiol implant plus water injection; empty implant plus progesterone injection; empty implant plus water injection. Blood samples were taken from animals in the morning and afternoon before and after various treatments. In animals not receiving estradiol, all short-day and some long-day females showed low morning values of LH and surge values in the afternoon. Estrogen suppressed morning and enhanced afternoon values of LH in long-day animals, intensifying the surge. However, in the presence of estrogen, these LH surges eventually diminished in both photoperiods. Progesterone hastened the loss of the LH surge. Hamsters did not consistently demonstrate FSH surges until treated with estrogen. Serum FSH in untreated hamsters was much higher in long-day than in short-day animals. Treatment with both steroids maximally suppressed morning and afternoon FSH levels in all hamsters. Thus, in the absence of estrogen, circadian expression of LH surges always occurs in short days; FSH surges sometimes occur. In all animals showing spontaneous or estrogen-induced surges, estrogen eventually leads to inhibition of gonadotropin secretion, particularly in the presence of progesterone.(ABSTRACT TRUNCATED AT 250 WORDS)