Curlews Numenius arquata use 2 methods to capture ragworms Nereis diversicolor they search for worms at the surface, which are taken with a single peck (called Npeck) and/or they search for visual cues, such as burrow entrances, and probe deep to extract the worm from the burrow (called Nprobe) Itis argued that male curlews select predominantly active worms, which are either grazing at the surface (Npeck) or are filter feeding within their burrows (Nprobe and sometimes Npeck). Profitability (mg S-' handling) increases with worm length and is greater for NPeck, because of its shorter handling time, than for NP,,&. Worms of both prey types I ca 6 cm are ignored, since profitability is below the intake rate (mg S-' feeding). During low water there is a shift from Nprobe to Npeck, which can be explained by changes in feeding behaviour of the worms. Nprobe, common in summer, disappears during autumn due to the increase of burrow depth (with the corresponding decrease of worms accessible as Nprobe), and to a reduction in filter feeding of the worm, which means that fewer traces are visible. Curlew search rate increases if, relatively, many prey are of the Nperk type. There appears to be a tradeoff between search rate and probab~lity of detection of conspicuous Npeck and cryptic Nprobe. NO evidence was found that curlews ignore one prey type and selectively search for the other We conclude that (1) unprofitable prey sizes are ignored, (2) the greater part of the profitable prey are not available. being hidden in the substrate and n~ostly not accessible even when detectable, (3) curlews continuously readjust their capture technique to changes in the available food supply.