In order to evaluate the relationship of PGE2 to hypercalcemia in cancer patients, 101 patients were screened with a radioimmunoassay for plasma prostaglandin E2 (PGE2) (NL less than 100 pg/ml). Of the 101 patients, 31 were hypercalcemia. Mean PGE2 (+/- SEM) of the 31 patients was 199 +/- 36 pg/ml. Among the 70 normocalcemic patients, mean +/- SEM PGE2 was 85 +/- 12 pg/ml (range = less than 25--225 pg/ml) (P less than 0.001). Seventeen hypercalcemic patients were initially treated with saline and furosemide, then were prospectively screened for serum parathormone (iPTH) and PGE2. Fourteen of 17 patients were then treated empirically with indomethacin (25 mg b.i.d.) for 72 hours and the PGE2 assay was repeated. Prior to therapy with indomethacin (mean +/- SEM), Ca++ = 12.2 +/- 1.5 mg/dl (NL 8.4--10.6 mg/dl), PGE2 = 87.1 +/- 36.8 pg/ml, (range = less than 25--209 pg/ml), and iPTH = 406 +/- 266 pg/ml (NL less than 400 pg/ml) (range = less than 100--825 pg/ml). PGE2 was elevated before treatment in 6/14 patients (breast, colon, renal, lung, neck tumors, and myeloma). Following treatment with indomethacin, PGE2 and calcium fell to normal levels in three patients (breast, colon, renal carcinomas). These results suggest: (1) A bimodal distribution of PGEs exists in hypercalcemic cancer patients. (2). There was some evidence of lack of whole molecule iPTH suppression in these patients. (3) Multiple stimuli of calcium mobilization may play an important etiologic role in a few hyercalcemic cancer patients and may explain the failure of indomethacin to control serum Ca++ in some patients with elevated PGE2.