Most ovarian tumors are cystic structures containing variable amounts of fluid. Several studies of ovarian cyst fluid focus on one specific metabolite using conventional assay systems. We examined the potential of 1 H-nuclear magnetic resonance spectroscopy in evaluation of the overall metabolic composition of cyst fluid from different ovarian tumors. Ovarian cyst fluid samples obtained from 40 patients with a primary ovarian tumor (12 malignant and 28 benign) were examined. After deproteinization and pD standardization, we performed 1 H-NMR spectroscopy on a 600 MHz instrument. With 1 H-NMR spectroscopy we found detectable concentrations of 36 metabolites with high intersample variation. A number of unassigned resonances as well as unexpected metabolites were found. We introduce an overall inventory of the low-molecular-weight metabolites in ovarian cyst fluid with corresponding resonances. Significant differences in concentration (p `0.01) were found for several metabolites (including an unknown metabolite) between malignant and benign ovarian cysts. Furthermore, higher concentrations in malignant-and lower in benign fluids were found compared to normal serum values, indicating local cyst wall metabolic processes in case of malignant transformation. We conclude that 1 H-nuclear magnetic resonance spectroscopy can give an overview of low-molecular-weight proton-containing metabolities present in ovarian cyst fluid samples. The metabolic composition of cyst fluid differs significantly between benign and malignant ovarian tumors. Furthermore, differences between benign subgroups possibly related to histopathological behaviour can be detected. The presence of N-acetyl aspartic acid and 5-oxoproline exclusively in serous cystadenoma samples is remarkable. Future studies will concentrate on these findings and explore the possibilities of extrapolating information from the in vitro studies to in vivo practice, in which metabolic differences between malignant and benign subtypes can be of great importance in a pre-operative phase.