Volume localization of magnetic resonance signals was achieved by using the regional susceptibility differences produced by superparamagnetic iron oxide particles. In vitro experiments demonstrated a direct linear relationship between the concentration of particulate iron and phosphorus-31 chemical shift or line broadening. In vivo experiments indicated that an intravenous dose of 5-10 mg of iron per kilogram of body weight suppressed P-31 signal from normal liver in healthy rats. In rats with hepatic implants of mammary adenocarcinoma, superparamagnetic iron oxide particles suppressed detectable P-31 or hydrogen-1 signal arising from healthy liver tissue, but not that from tumor. Signal due to surface tissues, which affect surface-coil spectra, could be selectively suppressed with a film-based application of particles to the abdominal wall. Thus, P-31 spectra from simulated or actual lesions could be selectively detected after chemically suppressing signals from neighboring or surrounding tissue.