Since David Lack first proposed that birds rear as many young as they can nourish, food limitation has been accepted as the primary explanation for variation in clutch size and other life-history traits in birds. The importance of food limitation in life-history variation, however, was recently questioned on theoretical grounds. Here, we show that clutch size differences between two populations of a neotropical thrush were contrary to expectations under Lack's food limitation hypothesis. Larger clutch sizes were found in a population with higher nestling starvation rate (i.e. greater food limitation). We experimentally equalized clutches between populations to verify this difference in food limitation. Our experiment confirmed greater food limitation in the population with larger mean clutch size. In addition, incubation bout length and nestling growth rate were also contrary to predictions of food limitation theory. Our results demonstrate the inability of food limitation to explain differences in several life-history traits: clutch size, incubation behaviour, parental feeding rate and nestling growth rate. These life-history traits were better explained by inter-population differences in nest predation rates. Food limitation may be less important to life history evolution in birds than suggested by traditional theory. ?? 2005 The Royal Society.