Sclerotherapy and sham-sclerotherapy were compared in male alcoholic patients with cirrhosis and bleeding esophageal varices. The prospective, single-blind, randomized clinical trial of 5 yr duration entered 253 male alcoholic patients at 12 Veterans Affairs medical centers. Patients were either actively bleeding from esophageal varices at randomization or they had a history of such bleeding. They were treated by endoscopy with sham-sclerotherapy or endoscopy with sclerotherapy at randomization, 4 to 6 days, 9 to 11 days, 1 mo, 3 mo and every 3 mo for 2 yr and followed for the remainder of the study period. Of the patients randomized, 131 were assigned to sham-therapy and 122 were assigned to sclerotherapy. At entry the two patient groups were comparable. The upper gastrointestinal rebleeding rates during the study were 101 and 66 per 100 person years of follow-up in sham-therapy and sclerotherapy, respectively (rate ratio, 1.54; 95% confidence interval, 1.06 to 2.22; p = 0.01). A significantly higher number of episodes were attributable to esophageal varices in the sham-therapy group (112 vs. 52; p = 0.005). Seventy-four sham-therapy patients (56%) and 77 sclerotherapy patients (63%) died (p = 0.54; relative risk, 0.91; 95% confidence interval, 0.66 to 1.25). The mean transfusion requirement was higher in the sham-therapy group (16.0 units vs. 9.4 units; p = 0.002) and more patients in this group required shunt surgery (18 vs. 5; p = 0.005). We conclude that sclerotherapy reduces the rate of rebleeding from varices, the transfusion requirement, and the need for shunt surgery, but has no long-term effect on survival.