Depletion of soil invertebrate availabilities by starling foraging during the breeding season was investigated by using exclosures to create areas where the birds could not feed. A field 'preferred' by the wild birds was compared with a 'non-preferred' field. Prey biomass was measured from soil cores and the foraging success of individuals feeding in experimental enclosures gave a measure of prey availability. Foraging by these birds resulted in significant short-term depletion after a few hours of foraging. Trials conducted after a 24-h interval showed that this depletion was due to removal of prey by the birds rather than invertebrate movements. Comparison with exclosure sites at the end of the chick-feeding period showed that foraging by the wild birds caused significant long-term depletion. This depletion was greatest in the 'preferred' field, which offered a higher availability of leatherjackets Tipula paludosa. The 'non-preferred' field yielded lower capture rates, despite a higher earthworm biomass in the soil cores. These findings suggest that soil core analysis alone is insufficient to predict the foraging success of starlings and other birds feeding on soil invertebrates. The significant reduction in prey availability to parent starlings is discussed in relation to the number of parental foraging trips made to the two fields.