Some b-blockers are efficiently removed from the circulation by hemodialysis ("high dialyzability") whereas others are not ("low dialyzability"). This characteristic may influence the effectiveness of the b-blockers among patients receiving long-term hemodialysis. To determine whether new use of a highdialyzability b-blocker compared with a low-dialyzability b-blocker associates with a higher rate of mortality in patients older than age 66 years receiving long-term hemodialysis, we conducted a propensity-matched population-based retrospective cohort study using the linked healthcare databases of Ontario, Canada. The high-dialyzability group (n=3294) included patients initiating atenolol, acebutolol, or metoprolol. The low-dialyzability group (n=3294) included patients initiating bisoprolol or propranolol. Initiation of a highversus low-dialyzability b-blocker was associated with a higher risk of death in the following 180 days (relative risk, 1.4; 95% confidence interval, 1.1 to 1.8; P,0.01). Supporting this finding, we repeated the primary analysis in a cohort of patients not receiving hemodialysis and found no significant association between dialyzability and the risk of death (relative risk, 1.0; 95% confidence interval, 0.9 to 1.3; P=0.71). b-Blocker exposure was not randomly allocated in this study, so a causal relationship between dialyzability and mortality cannot be determined. However, our findings should raise awareness of this potentially important drug characteristic and prompt further study.