Although beta blockers have had significant impact in the treatment of portal hypertension, the question of how long they should be continued for prevention of variceal hemorrhage remains unknown. Prospective studies on beta blockers to prevent variceal hemorrhage lack long-term follow-up, and indefinite administration of beta blockers for primary prevention of variceal bleeding has become standard practice. The aim of this study was to determine the outcomes of patients in whom beta blocker therapy was discontinued. Patients completing a prospective, randomized, double-blind, placebo-controlled trial of propranolol for the primary prevention of variceal hemorrhage were tapered off of propranolol and placebo and followed prospectively for subsequent events. Of the 49 patients in the follow-up study (25 former propranolol, 24 former placebo), 9 experienced variceal hemorrhage (6 former propranolol, 3 former placebo). Following withdrawal of propranolol, the freedom from variceal bleeding was not significantly different between these 2 groups of patients, suggesting that the protective effect of propranolol against variceal hemorrhage, noted previously, was no longer present. Seventeen patients died (12 former propranolol, 5 former placebo) during the follow-up study. Cumulative survival was longer in the placebo group. These trends for EVH and survival were opposite to those observed in the original study population while patients were taking medication. When propranolol is withdrawn, the risk of variceal hemorrhage returns to what would be expected in an untreated population. Patients who discontinue beta blockers experience increased mortality compared with an untreated population. These observations support the current practice of indefinite prophylactic therapy.