To determine the impact of indomethacin on the course of uremic pericarditis we performed a prospective, double blind study in which 24 patients with endstage chronic renal failure and pericarditis randomly received indomethacin, 25 mg four times daily, (11 patients) or a placebo (13 patients) for a 3-week period. All patients received peritoneal or hemodialysis treatment concurrently with the study drug. In contrast to the placebo, indomethacin produced an immediate and sustained reduction of fever in all but one patient. On the other hand, indomethacin had no effect on the duration of chest pain (mean days +/- SE: placebo 1.4 +/- 0.6, indomethacin 5.5 +/- 3.3), duration of pericardial friction rub (placebo 10.3 +/- 1.7, indomethacin 16.0 +/- 3.8), or on the amount of pericardial effusion. Further, indomethacin did not diminish the need for invasive surgical procedures for relief of tamponade (three of 13 placebo patients, two of 11 indomethacin patients) or result in decreased mortality rate. Death (not due to pericarditis) occurred in two patients treated with indomethacin and one patient who received the placebo. In our patients pericarditis encompassed a wide spectrum ranging from a mild illness of several days duration to a painful and debilitating disease lasting weeks and requiring surgical intervention. Although the size of our population prohibits definitive conclusions, it would appear that, except for fever, the manifestations and natural history of this illness are unaffected by indomethacin.