The Ideal Free Distribution was developed to predict the distribution of organisms at a habitat level. The theory of the Ideal Free Distribution assumes that travel between resource sites has a negligible affect on the distribution of organisms. In this experiment we tested whether the Ideal Free Distribution, and in particular its prediction of habitat matching, is robust to violations of this assumption. In an experiment with free-living ducks we manipulated the distance between two food sites. We used two conditions, one with 16 m between the two food sites and another with 45 m between the resource sites. We found that the distribution of organisms became less extreme with increased travel distance. This result is probably due to two effects: that travel distance caused a decrease in the ducks' ability to discriminate between the sites' profitabilities and by a decrease in the number of ducks travelling between the resource sites with increased distance. The decreased number of ducks travelling alone can explain only a relatively small amount of the change in the distribution. The decrease in discriminability may be due to either (or both) the increased distance causing a decrease in the foragers' ability to visually judge the relative profitabilities of the sites or by a decrease in switching rate associated with travel distance (if physical sampling of a site is needed to gather information). Because even a minor change in travel distance can cause a significant change in the distribution of foraging organisms, caution is urged about making extrapolations from experiments at a small spatial scale to the habitat level.