Liver biopsy remains the gold standard for characterizing diffuse liver disease and is associated with significant morbidity and, rarely, mortality. Our aim was to investigate whether a noninvasive technique, in vivo phosphorus 31 ( 3 1 P)-magnetic resonance spectroscopy (MRS), could be used to assess the severity of hepatitis C virus (HCV)-related liver disease. Fifteen healthy controls and 48 patients with biopsy-proven HCV-related liver disease were studied prospectively. Based on their histologic fibrosis (F) and necroinflammatory (NI) scores, patients were divided into mild hepatitis (F =< 2/6, NI =< 3/18), moderate/severe hepatitis (3 =< F < 6 or NI =< 4/18), and cirrhosis (F = 6/6). Hepatic 3 1 P MR spectra were obtained using a 1.5-T spectroscopy system. Quantitation of the 3 1 P signals was performed in the time domain using the Advanced MAgnetic RESonance algorithm. There was a monotonic increase in the mean +/- 1 standard error phosphomonoester (PME) to phosphodiester (PDE) ratios for the control, mild disease, moderate disease, and cirrhosis groups: 0.15 +/- 0.01, 0.18 +/- 0.02, 0.25 +/- 0.02, 0.38 +/- 0.04, respectively (ANOVA, P < .001). An 80% sensitivity and specificity was achieved when using a PME/PDE ratio less than or equal to 0.2 to denote mild hepatitis and a corresponding ratio greater than or equal to 0.3 to denote cirrhosis. No other significant spectral changes were observed. In conclusion, 3 1 P MRS can separate mild from moderate disease and these 2 groups from cirrhosis. The ability to differentiate these populations of patients has therapeutic implications and 3 1 P MRS, in some situations, would not only complement a liver biopsy but could replace it and be of particular value in assessing disease progression. (Hepatology 2003;37:788-794.)