The rapid increase in body mass that often occurs following creatine (Cr) supplementation is believed to be due to intracellular water retention. The purpose of this study was to determine whether Cr consumption alters the magnetic resonance (MR) transverse relaxation (T2) distribution of skeletal muscle. Transverse relaxation can be used to model water compartments within a cell or tissue. In this double-blind study, subjects were asked to supplement their normal diet with creatine monohydrate (20 g day-1 for 5 days) mixed with a grape drink (Creatine group, n = 7), or the grape drink alone (Placebo group, n = 8). Phosphorous MR spectroscopy was used to determine the effectiveness of the supplementation protocol. Subjects that responded to the Cr supplementation (i.e. showed a > 5 % increase in the ratio of the levels of phosphocreatine (PCr) and ATP) were placed in the Creatine group. Both proton MR imaging and spectroscopy were used to acquire T2 data, at 1.89 T, from the flexor digitorum profundus muscle of each subject before and after supplementation. Following the supplementation period, the Creatine group showed a gain in body mass (1.2 ± 0.8 kg, P < 0.05, mean ± S.D.), and an increase in PCr/ATP ratio (23.8 ± 16.4 %, P < 0.001). Neither group showed any changes in intracellular pH or T2 calculated from MR images. However, the spectroscopy data revealed at least three components (> 5 ms) at approximately 20, 40 and 125 ms in both groups. Only in the Creatine group was there an increase in the apparent proton concentration of the two shorter components combined (+5.0 ± 4.7 %, P < 0.05). According to the cellular water compartment model, the changes observed in the shorter T2 components are consistent with an increase in intracellular water. Experimental Physiology (2002) 87.3, 383-389.