Techniques for the separation of small quantities of Cr from carbonate material and for the analysis of stable Cr isotopes in carbonates by MC-ICP-MS are presented in this study. In comparison with previously published methods, we have developed a one-step Cr separation procedure that is relatively simple, and has a low blank (0.12–0.20 ng). Moreover, careful optimisation of the desolvating sample introduction system allows a significant increase in the sensitivity of our MC-ICP-MS technique compared to previous studies. Instrumental mass bias effects and fractionation of Cr isotopes during Cr separation are corrected using a carefully optimised 50Cr–54Cr double-spike method. Novel numerical simulations demonstrate that the effects of potential isobaric interferences from Ti, Fe and V are negligible, even if they are isotopically fractionated. Small deviations in the δ53Cr value of the NBS 979 standard between different analytical sessions are due to small deviations from exponential mass fractionation behaviour. The long-term reproducibility of δ53Cr for a spiked NBS 979 Cr isotope reference material is ±0.031‰ (2 S.D., n = 147). Analyses of carbonates reveal that they have δ53Cr values of 0.747 to 1.994‰, distinctly heavier than continental crust and the terrestrial mantle. The carbonates record Cr isotopic fractionation that may be used to understand redox reactions in the oceans. Although this study focuses on carbonate samples, our mass spectrometry technique can be applied to the analysis of any samples with low levels of Cr, including river waters and seawater.