Use of stable isotope ratios to trace pathways of organic matter among consumers requires knowledge of the isotopic shift between diet and consumer. Variation in trophic shift among consumers can be substantial. For data from the published literature and supplementary original data (excluding fluid-feeding consumers), the mean isotopic shift for C was +0.5±0.13‰ rather than 0.0‰, as commonly assumed. The shift for C was higher for consumers analyzed as muscle (+1.3±0.30‰) than for consumers analyzed whole (+0.3±0.14‰). Among consumers analyzed whole, the trophic shift for C was lower for consumers acidified prior to analysis (−0.2±0.21‰) than for unacidified samples (+0.5±0.17‰). For N, trophic shift was lower for consumers raised on invertebrate diets (+1.4±0.21‰) than for consumers raised on other high-protein diets (+3.3±0.26‰) and was intermediate for consumers raised on plant and algal diets (+2.2±0.30‰). The trophic shift for S differed between high-protein (+2.0±0.65‰) and low-protein diets (-0.5±0.56‰). Thus, methods of analysis and dietary differences can affect trophic shift for consumers; the utility of stable isotope methods can be improved if this information is incorporated into studies of trophic relationships. Although few studies of stable isotope ratios have considered variation in the trophic shift, such variation is important because small errors in estimates of trophic shift can result in large errors in estimates of the contribution of sources to consumers or in estimates of trophic position.