Accurate determination of feeding habits using only conventional dietary analysis is usually difficult, due to both biases caused by prey digestibility differences and to the discrete aspect of the sampling. However, combining conventional methods with stable isotope analysis provides an integrated view of the assimilated diet. Here, we measured stable isotope ratios of carbon (δ13C), nitrogen (δ15N) and sulphur (δ34S) in feathers of yellow-legged gull Larus michahellis nestlings from 4 colonies along the western Mediterranean coast, where the gulls consume different proportions of marine, terrestrial and garbage resources. We collected and analysed chicks' regurgitates in each colony, thereby determining the isotopic signatures in a significant sample of prey. By applying a mixing model to our isotopic values, we compared the information provided by the 2 methodologies. According to direct prey sampling and the particular conditions of the breeding areas, populations with more enriched δ34S values consumed a diet richer in marine prey, with lower δ15N signatures indicating garbage consumption. The main pattern resulting from the mixing model agreed with direct diet sampling. However, estimated proportions for small-sized prey were controversial, which indicates that small, soft prey items might be underestimated by regurgitate analysis. We conclude that stable isotope signatures and the use of mixing models are useful tools for the rapid assessment of feeding ecology in certain populations. Solving analytical biases should be considered in future feeding studies, thereby saving time and minimising bird disturbance by using an isotopic methodology.