The effects of postoperative subconjunctival 5-fluorouracil (5-FU) injections on the surgical outcome of trabeculectomy in the Japanese were studied in a total of 362 eyes of 263 glaucoma patients with the postoperative follow-up of 1-9 years. The probability of successful intraocular pressure (IOP) control at the 5-year point of the continuing follow-up was calculated by the life-table method of Kaplan-Meier. The eyes that had undergone trabeculectomy with postoperative injections of 5-FU (5-FU group) included 165 eyes with primary open-angle glaucoma (POAG), 38 eyes with secondary glaucoma (SG) and 60 eyes with refractory glaucoma. The eyes that had undergone trabeculectomy without postoperative 5-FU (control group) included 46 eyes with POAG, 31 eyes with SG and 22 eyes with refractory glaucoma. The surgical techniques and postoperative care were virtually identical between the two groups, except that the control group did not receive the postoperative 5-FU. The total amount of 5-FU given to each case averaged 36.5, 36.0 and 49.5 mg for POAG, SG and refractory glaucoma, respectively. At the 5-year follow-up point, the probability of maintaining IOP control below 21 mmHg with or without medication in the 5-FU group was 92.5% for POAG, 87.4% for SG and 57.5% for refractory glaucoma; while without medication the probability was 58.2, 54.8 and 27.8%, respectively. The probability of IOP control below 16 mmHg with or without medication in the 5-FU group was 77.9% for POAG, 66.8% for SG and 26.9% for refractory glaucoma; without medication the probability was 55.2, 49.8 and 24.5%, respectively. In the control group, the corresponding probability for each type of glaucoma was much lower than that in the 5-FU group, and the probability for IOP control below 16 mmHG was 0.12% at the 5-year follow-up point. In the 5-FU group, the mean postoperative IOP in successfully treated eyes remained at about 12 mmHg throughout the first 5 years of the follow-up period. There was no significant difference in incidence of postoperative complications between the 5-FU and control groups, except that corneal epithelial defect was noted in 38.8% of the eyes in the 5-FU group. No late complications related to the use of postoperative 5-FU were observed.(ABSTRACT TRUNCATED AT 400 WORDS)