Fifty cirrhotic patients with esophageal varices underwent endoscopic treatment in a prospective randomized trial carried out to compare variceal ligation with injection sclerotherapy, with respect to safety, efficacy, and complications in the initial session. Twenty-five patients each were treated using endoscopic variceal ligation or endoscopic injection sclerotherapy in the initial session and then only endoscopic injection sclerotherapy for all following sessions. Ligations numbered 4.1 at the initial session. The total number of sessions of endoscopic injection sclerotherapy was 4.0 +/- 1.3 in the endoscopic injection sclerotherapy group compared with 3.6 +/- 1.4 in the endoscopic variceal ligation group. The total volume of the sclerosant used by the end of all the sessions was 23.9 +/- 10.3 ml in the endoscopic variceal ligation group and 39.0 +/- 11.5 ml in the endoscopic injection sclerotherapy group (p < 0.0001). In all patients, complete eradication of the varices was achieved, and at 7 to 15 months follow-up no evidence of recurrence was seen. Adverse effects such as pyrexia, chest pain, and pleural effusion after the initial session were significantly lower in the endoscopic variceal ligation group (p < 0.05), and total bilirubin, serum creatinine, lactate dehydrogenase, and urinary beta 2-microglobulin were significantly increased and PaO2 decreased after endoscopic injection sclerotherapy (p < 0.05). This study shows that endoscopic variceal ligation significantly decreased the adverse effects associated with endoscopic injection sclerotherapy in the initial session, and it is recommended as an alternative to sclerotherapy.