1. Lipids have more negative delta(13)C values relative to other major biochemical compounds in plant and animal tissues. Although variable lipid content in biological tissues alters results and conclusions of delta(13)C analyses in aquatic food web and migration studies, no standard correction protocol exists. 2. We compared chemical extraction and mathematical correction methods for freshwater and marine fishes and aquatic invertebrates to better understand impacts of correction approaches on carbon (delta(13)C) and nitrogen (delta(15)N) stable isotope data. 3. Fish and aquatic invertebrate tissue delta(13)C values increased significantly following extraction for almost all species and tissue types relative to nonextracted samples. In contrast, delta(15)N was affected for muscle and whole body samples from only a few freshwater and marine species and had a limited effect for the entire data set. 4. Lipid normalization models, using C : N as a proxy for lipid content, predicted lipid-corrected delta(13)C for paired data sets more closely with parameters specific to the tissue type and species to which they were applied. 5. We present species- and tissue-specific models based on bulk C : N as a reliable alternative to chemical extraction corrections. By analysing a subset of samples before and after lipid extraction, models can be applied to the species and tissues of interest that will improve estimates of dietary sources using stable isotopes.