We investigated experimentally the seasonal role of food supply in brood survival of the European coot Fulica atra. For two breeding seasons, individual pairs were offered supplemental food the first ten days after their young hatched. Under natural conditions this period was largely responsible for the seasonal variation in brood survival. Our experiment tested three hypotheses: (1) food supply is not involved in breeding success at any date (other factors hypothesis), (2) food supply limits success independently of date (elevation hypothesis), and (3) food supply affects success seasonally and is responsible for the natural seasonal trend in success (date hypothesis). Experimental pairs with supplemental food raised heavier, larger and more chicks than control pairs. Consistent with the date hypothesis, food supplementation abolished seasonal variation in chick survival. Chick growth under supplemental food was in agreement with the elevation hypothesis. This discrepancy was probably due to the limited supply of additional food. We conclude that seasonal variation in offspring production was in essence the result of seasonal variation in food availability.