Forty-seven patients with esophageal variceal bleeding were randomly allocated to undergo sclerotherapy on a 3 weekly schedule with either 5% ethanolamine oleate (23 patients) or absolute alcohol (24 patients), in an attempt to compare the efficacy and safety of the two sclerosants. Sclerotherapy with absolute alcohol eradicated esophageal varices significantly earlier compared with ethanolamine oleate (12.9 +/- 5.2 vs 22.3 +/- 8.2 wk, respectively, p less than 0.001). The mean number of injection courses and the mean amount of sclerosant required for variceal obliteration was also significantly (p less than 0.001) less in the alcohol-injected group. Although the total number of rebleeding episodes were significantly (p less than 0.05) less in the alcohol-injected group, the frequency of rebleeding was not significantly different between the two groups (20.8% vs 30.4%, respectively, p greater than 0.05). Two (8.1%) patients died due to rebleeding in the ethanolamine-injected group, whereas in the alcohol group, none died. There was no significant difference in the frequency of complications with the two sclerosants. Besides the relative ease of rapid injection due to its aqueous nature, alcohol is readily available and relatively economical (total cost of sclerosant per patient; alcohol US $0.50, ethanolamine US $60). In conclusion, absolute alcohol appears to be a useful alternative to 5% ethanolamine oleate as a variceal sclerosant.