The great diversity of colour patterns in birds remains largely unexplained (Burtt 1986, Butcher and Rohwer 1989). Experimental and comparative studies have been used to identify some adaptive correlates, including intra-specific communication (Hill 1990, 1994, Slagsvold and Saetre 1991, Saetre and Slagsvold 1992, Marchetti 1993), camouflage (Gaston 1974, Burtt 1986), abrasion resistance (Burtt 1986), inter-specific signalling (Baker and Parker 1979, Gbtmark 1993, 1995, Saetre et al. 1993) and flushing out prey (Jablonski 1996). While we might expect plumage patterns to represent a compromise between all these and other selection pressures, two hypotheses have been recently proposed to explain the diversity of colour patterns within the genus Phylloscopus (Old World leaf warblers). Marchetti (1993) ascribes a role for intra-specific communication (i.e., signalling to potential mates or competitors) in the evolution of colour patterns, whereas Jablonski (1996) suggests that conspicuous coloration might be important for flushing prey. In this note we assess the role of prey-flushing in the Phylloscopus warblers using quantitative data on foraging and colour patterns. Jablonski (1996) proposed that wing, rump and tail patches could be important in prey-flushing in the Phylloscopus warblers. To support this idea he used qualitative reports on foraging behaviour culled from the literature, and a cross-species comparative study to demonstrate an association of conspicuousness with the use of the wings during foraging. However, Burtt (1986) in a comparative study of the New World warblers also found an association between plumage brightness and foraging movements, but interpreted it as being a consequence of intra-specific communication. For example, species which flycatch are prominently displaying their wings, and therefore the wings are suitable locations to place patterns used in intra-specific communication. From Jablonski's prey-flushing hypothesis we predict positive correlations between plumage brightness and only those flying movements that lead to chases after flushed prey, but from Burtt's intra-specific communication hypothesis we expect a positive correlation between plumage brightness and the number of all types of flying movements performed by the bird when feeding. There are therefore two purposes to this paper. First, we demonstrate a correlation between plumage brightness and foraging among the Phylloscopus warblers using more quantitative data than were used by Jablonski (1996). Second, we use our results to compare the predictions of the intra-specific communication and prey-flushing hypotheses. Phylloscopus species differ in the number of colour patches they have. In all species, colour patches occur as bright, unpigmented areas on the tips of certain feather tracts of the body. Some species are essentially unpatterned and have uniform dark upperparts except for an eye-stripe. Others have one or two patches on the wings, a crown stripe, a rump patch and/or white tail patches in nested combinations, so that, for example, every species with crown or rump patches always has a wing patch (see Marchetti 1993, Price and Pavelka 1996). We used three different measures of plumage brightness for a species: 1. the total number of colour patches, varying from 0-5 (Marchetti 1993). 2. the total brightness (% reflectance relative to a standard, as measured by reflectance spectroradiometry), of the colour patch at the tip of the feather taken from the tract of feathers which form the largest wing patch on those species which have a wing patch (Marchetti 1993). 3. the size of the unpigmented patch on this same feather (area without melanin; Price and Pavelka 1996). The three variables are correlated (Fig. 1, Table 1). Quantitative data on foraging are available for eight Phylloscopus species from Kashmir, India (Price 1991). Feeding methods were classified into four main types: standpick (a prey capture not involving flight), flypick (prey resting on a leaf or branch, observed while hopping, but taken in flight), flycatch (capture of prey, which was nearly always flushed), and hoverpick (prey