Stable carbon isotope analysis of animal liver and muscle has become a widespread tool for investigating dietary ecology. Nonetheless, stable carbon isotope turnover of these tissues has not been studied in large mammals except with isotopically labelled tracer methodologies, which do not produce carbon half-lives analogous to those derived from naturalistic diet-switch experiments. To address this gap, we studied turnover of carbon isotopes in the liver, muscle, and breath CO2 of alpacas (Lama pacos) by switching them from a C3 grass diet to an isonitrogenous C4 grass diet. Breath samples as well as liver and muscle biopsies were collected and analyzed for up to 72 days to monitor the incorporation of the C4-derived carbon. The data suggest half-lives of 2.8, 37.3, and 178.7 days for alpaca breath CO2, liver, and muscle, respectively. Alpaca liver and muscle carbon half-lives are about 6 times longer than those of gerbils, which is about what would be expected given their size. In contrast, breath CO2 turnover does not scale readily with body mass. We also note that the breath CO2 and liver data are better described using a multiple-pool exponential decay model than a single-pool model. Copyright © 2006 John Wiley & Sons, Ltd.