Recent behavioural and molecular studies have shown that in most monogamous bird species extra–pair copulations and fertilizations outside the pair bond occur routinely. The consequences of female extra–pair behaviour might comprise effects on important life–history traits, such as the extent of male parental care. In this study we test the assumption that, within a species, females' options for extra–pair mating depend on female quality and the environments that females occupy. This‘constrained female hypothesis’ predicts that females in good environments or high–quality females are able to resist males' control efforts better than females in poor environments or low–quality females. We test the idea in the socially monogamous serin. We found that the likelihood of extra–pair paternity is significantly higher in territories with high availability of food. There was a negative relationship between environmental quality (food availability) and paternity both in natural and in experimentally manipulated habitats. Male feeding rates were negatively related to food availability and positively related to paternity. These data and the additional result that in better environments all of a females' offspring were sired by one extra–pair male provide support for Gowaty's ‘constrained female hypothesis’.