RATIONALE: Dorsal white muscle is the standard tissue analysed in fish trophic studies using stable isotope analyses. However, sampling white muscle often implies the sacrifice of fish. Thus, we examined whether the non-lethal sampling of fin tissue can substitute muscle sampling in food web studies. METHODS: Analysing muscle and fin d15N and d13C values of 466 European freshwater fish (14 species) with an elemental analyser coupled with an isotope ratio mass spectrometer, we compared the isotope values of the two tissues. Correlations between fin and muscle isotope ratios were examined for all fish together and specifically for 12 species. We further proposed four methods of assessing muscle from fin isotope ratios and estimated the errors made using these muscle surrogates. RESULTS: Despite significant differences between isotope values of the two tissues, fin and muscle isotopic signals are strongly correlated. Muscle values, estimated with raw fin isotope ratios (1st method), induce an error of ca. 1% for both isotopes. In comparison, specific (2nd method) or general (3rd method) correlations provide meaningful corrections of fin isotope ratios (errors <0.6%). On the other hand, relationships, established for Australian tropical fish, only give poor muscle estimates (errors >0.8%). CONCLUSIONS: There is little chance that a global model can be created. However, the 2nd and 3rd methods of estimating muscle values from fin isotope ratios should provide an acceptable level of error for the studies of European freshwater food web. We thus recommend that future studies use fin tissue as a non-lethal surrogate for muscle.