Stable isotopes are increasingly used in ecology to study the diets, trophic position, and migratory patterns of wildlife including herpetofauna. When using stable isotopes, it is important to consider which tissues can or should be sampled, and how selecting tissues may affect the inferences drawn from stable isotope data. Amphibians offer fewer tissues than other larger organisms that can be harvested in sufficient quantity without killing the animal; however, many salamanders have tails that readily autotomize and regenerate. We used three species of plethodontid salamander (Plethodon cinereus, P. metcalfi, and Desmognathus quadramaculatus) to determine whether distal tail tissue had carbon and nitrogen stable isotope values comparable to commonly used tissues (liver and whole carcass [minus the liver and gonads]) that must be collected lethally. We found that variation in carbon values (δ13C) within and among tissues was negatively correlated with C:N (an indication of high lipid content). Nitrogen (δ15N) and Carbon (δ13C) values (once adjusted for C:N) of tail tissue were positively correlated with values for carcass tissue, and tail δ15N values were positively correlated and biased with δ15N values for liver tissue. Adjusted δ13C values for tail tissue were unbiased and had weaker positive correlations with δ13C values for liver tissue compared to whole carcass. The weak correlation between adjusted δ13C values of tail and liver tissues may reflect differences in turnover rates between the two tissues. Our data show that, with calibration, non-lethal collection of tail clips is a suitable substitute to lethal tissue collection for measuring δ13C and δ15N of plethodontid salamanders.