Serum gonadotropin hormone (GTH) concentrations were determined by radioimmunoassay in goldfish Carassius auratus subjected to short photophase and cold or warm temperature, or long photophase and warm temperature for various lengths of time in winter and spring. The effects of an outdoor pond regime, phase-shifting of the photoperiod and feeding times, warm temperature pulses, pinealectomy, blinding, and melatonin treatment on the daily cycles in serum GTH concentrations were investigated. Ovarian condition was assessed by histology to allow correlation with GTH. The GTH cycles detected in the laboratory were similar to the cycles found in fish kept in an outdoor pond under similar environmental regimes. In addition to photoperiod and temperature, the length of the acclimation period had an effect on the patterns of the daily cycles in serum GTH concentrations. Timing of daily peaks in serum GTH concentrations was shifted by shifting of the photoperiod and feeding times. Photoperiod and feeding entrained the daily cycle in serum GTH when the onset of light and first feeding were 4 hours apart but not when they were 10 hours apart. A pulse of warm temperature promoted fluctuations in GTH when the warmth was imposed during the day, but not when imposed during the night. Pinealectomy and blinding suppressed fluctuations in serum GTH concentrations under long photophase and warm temperature, but pinealectomy promoted fluctuations under short photophase and warm temperature. Melatonin, a pineal indoleamine, lowered the GTH peak when administered in the morning; however, no effect was detected from the afternoon injections. In most experiments, abolition of the daily fluctuations in serum GTH concentrations was correlated with smaller ovaries or ovarian atresia. The daily cycle in serum GTH promoted ovarian growth more than steady low or high serum GTH concentrations.