The handicap hypothesis assumes that sexual ornaments impose a viability cost upon the bearers. There have been few empirical tests of this assumption. Previous studies show evidence for the cost of a tail ornament in male birds: a negative relationship between an experimentally increased tail ornament (long tail streamers) and efficiency at foraging for nestlings. However, it must be admitted, that the apparent impairing effect of an elongated tail could be a result of a decrease in male parental effort in response to an increase of female parental effort, which might have occurred in response to increased male attractiveness (differential allocation of female parental effort). In this study, the effect of differential parental expenditure was eliminated by lengthening the tail in female, rather than male, sand martins (Riparia riparia). Tail-elongated females decreased the rate at which they fed nestlings, and captured more but smaller insects. There was no simultaneous increase of feeding rate in the males that could explain the decrease of feeding rate in the females. These results confirm the existence of a cost of a tail ornament in birds feeding in flight, as is expressed in terms of impaired flight and foraging capacity.