Equidae are born in the spring. This is due to the fact that mares go through a period of anovulation from October to the end of April. For breeders, bringing forward the first ovulation presents an enormous advantage, as it increases the number of usable cycles in the mare and increases the actual age of the offspring at the time of sports performance or slaughter. In mares under our latitudes, the first ovulation can be moved to its earliest by using 14.5 h of daily light for 35 days, beginning around the winter solstice. This treatment is robust, as repeated accidents do not affect its outcome. The administration of melatonin to pony mares during winter ovarian inactivity, either daily in the evening or with a subcutaneous implant, suppresses the photostimulating effect of long days (LD). On the other hand, the use of this implant around the summer solstice brings forward the first ovulation, as in photostimulated mares, thereby confirming the existence of an annual endogenous rhythm. These implants inserted under the same conditions in non-pregnant mares do not induced ovarian inactivity. Therefore, the role of melatonin seems limited to the timing of winter inactivity. In thin but healthy mares subjected to the natural photoperiod, winter inactivity is systematic and long lasting. This inhibition, induced by the feed restriction, masks the effect of the photostimulation. On the other hand, 60% of fat mares have cycles all year long, and in the remaining 40%, the period of inactivity is very short, missing only one or two cycles. Increasing the feed in thin mares in autumn or winter seems to have no effect on the date of the first annual ovulation. Nutrition is the main cause of ovarian inactivity in winter, but it does not play a role in the resumption of activity in the spring. The mechanism inhibiting the cycles in autumn results from an interaction between short days and a negative energy balance. This nutrition-photoperiod interaction seems to be applicable to numerous species in temperate areas.