The present work describes methodology for the precise and accurate determination of chromium isotope ratios (50Cr/52Cr, 53Cr/52Cr, and 54Cr/52Cr) by multicollector inductively coupled plasma mass spectrometry (MC-ICP-MS). Mass discrimination and instrument drift are corrected for using a “standard–sample–standard” bracketing technique and normalization against NIST SRM 979 Cr(NO3)3·9H2O. The precisions (2s) of ratio measurements for 50Cr/52Cr, 53Cr/52Cr, and 54Cr/52Cr are ±0.11‰, ±0.06‰, and ±0.10‰ (110 ppm, 60 ppm and 100 ppm), respectively. For geological samples of relatively high concentration (300–17 000 ppm), Cr was separated from the matrix using columns of anion and cation exchange resins, providing a recovery of >95%. Inexplicably, for some geological samples, the aforementioned methodologies were not effective in reducing the concentrations of Fe and Ti to below 2% relative to the Cr concentration in the eluate. As a consequence, only 53Cr/52Cr was determined in such samples. Significant differences in isotope ratios were found among standard (NIST SRM 979), chemically pure metal chips of Cr (Alfa Aesar), a Late Proterozoic gabbro, a Cretaceous oil shale and natural products of combustion metamorphism of that oil shale. The igneous rocks and the pure metal have slightly negative δ53/52Cr values, the oil shale is +1.01‰, and the products of combustion metamorphism are all somewhat lighter, suggesting that the heavier isotopes of chromium were removed in the process of metamorphism.