The pathophysiological mechanisms to explain peritumoral edema have not been clarified. Multiple aspects of brain edema secondary to supratentorial meningiomas were prospectively investigated in a group of 29 patients who underwent surgery consecutively. Sixty-nine tumor samples were analyzed for prostanoid levels. Levels of 6-keto-PGF1 alpha, the stable metabolite of prostacycline, were found to correlate well with the extent of edema (r = 0.51, P < 0.01). The ratio, 6-keto-PGF1 alpha x PGE2/TXB2, was found to have the best correlation with edema index (extension/tumor volume) (r = 0.69, P < 0.005). A case of a hemangiopericytic meningioma with the largest edema extent within the study group also exhibited the highest level of 6-keto-PGF1 alpha (2420 pg/mg protein). Steroid treatment (dosage, duration of therapy, and their product) did not correlate with prostaglandin levels. These findings may explain the inconsistent clinical effects of steroids on meningioma-induced edema. Possible explanations for this phenomenon are discussed. Otherwise, histology, pathological features of tumor aggressiveness, or mechanical parameters, such as its volume, location, and insertion site, did not correlate well with edema parameters or with prostaglandin levels. Similarly, tumor water content, imaging parameters in computed tomography and magnetic resonance, and operative findings (including dissection plane, vascularity, and tumor firmness) did not correlate well with edema parameters. Although a direct cause-effect relationship between prostaglandins and peritumoral edema is not conclusively established, the circumstantial evidence of the ability of prostaglandins to induce vasogenic brain edema and the robust association with peritumoral edema is persuasive.(ABSTRACT TRUNCATED AT 250 WORDS)