How costly is provisioning chicks and how fast should birds fly to do it?These are the questions Kyle Elliott and Anthony Gaston sought to answer by measuring the flight speeds of seabirds in the field. Over the past 30 or so years, numerous models have been developed to estimate the power requirements for flapping flight in birds. The models predict a U-shaped relationship between power requirements for flight and flight speed. From these models, the flight speeds requiring least power (Vmp) and giving maximum range (Vmr) can be calculated. From such models,R. Å. Norberg predicted in 1981 that birds feeding chicks should fly faster than Vmr to maximise energy delivery to nestlings. Elliott and Gaston sought to test this hypothesis by making field measurements of flight speeds of Brünnich's guillemot and northern fulmar before and after their chicks had been hatched.Guillemot chicks fledge before completing growth, which has been suggested to be due to the inability of parents to maintain the cost of provisioning. By contrast, it is thought that fulmars are under time, rather than energy,constraints. Elliott and Gaston hypothesised that both species should fly faster during chick rearing than during incubation to maximise chick growth rates. By using infra-red distance-sensing binoculars, they were able to time birds of both species over a distance of 850 m on a flyway into the breeding colony. The velocities measured during inbound and outbound flights were assumed to be representative of those during entire foraging trips.Guillemots were found not to differ in flight speed between incoming and outgoing flights and before and after their chicks hatched. By contrast,fulmars flew faster after their chicks hatched, and incoming flights were faster than outbound ones. guillemots, it seems, fly close to Vmr while fulmars fly closer to Vmp. Elliott and Gaston propose a number of explanations for possible causes of quantitative disagreement with predicted flight speeds. Fulmars have a`flatter' power curve than guillemots, so it is metabolically less costly for them to adjust flight speed. Furthermore, patterns of weight loss, and hence prey-carrying capacity, differ between the two species. The fulmar may lose 15% of its mass when brooding compared with during incubation, compared with 5% in the guillemot. This mass change might lead to a predicted decrease in flight speed close to that observed. As the authors point out, however, their estimates of power are dependent upon the assumption that birds make their entire outbound and inbound flights at speeds similar to those recorded in their measurement area close to the colony.This paper demonstrates that quantifiable differences in flight speeds are observed in some species of bird in response to different demands of chick provisioning and demonstrates that field measurements can go some way to validating the predictions made concerning the tradeoffs between power requirements, range and velocity in seabirds.