An examination was made of the possibility that 100 to 200 ml of intraperitoneal 32% dextran 70 and/or 0.5% hydrocortisone sodium succinate (randomized independently with similar volumes of Ringer's lactate solution) might help to lessen the postoperative formation of adhesions among patients undergoing surgical procedures for peritubal adhesions (n = 76), endometriosis (n = 27), or midtubal occlusion (n = 61). Patients in the first two groups who were given intraperitoneal corticosteroids were also given systemic steroids. Nonparametric comparison of median adhesion scores at operation and at subsequent laparoscopy showed that there was a poorer outcome with dextran than when dextran was not used in every subgroup except one (repeat salpingolysis after previous operation for adhesions), including first operations for adnexal adhesions (Mann-Whitney U = 200, m = 23, n = 26; p less than 0.05). The probability was small (p beta less than 0.002) that an important beneficial effect of dextran was overlooked. Systemic corticosteroids were associated with a consistent trend toward improved outcome, especially in patients who initially had few or no adhesions, such as those operated on for endometriosis (U = 2, m = 7, n = 4; p less than 0.025), but among patients with tubal resections and anastomoses with adnexal adhesions the use of intraperitoneal hydrocortisone alone was associated with a worse outcome (U = 15, m = 12, n = 8; p less than 0.02). Life-table analysis of the accumulating probability of pregnancy showed that no significant difference resulted from adjunct use in any group. The conclusion is that no empiric basis supports the use of intraperitoneal 32% dextran 70 or 0.5% hydrocortisone in the attempt to prevent peritoneal adhesions, but further investigations on the systemic administration of corticosteroids to decrease the formation of adhesions would be useful.