OBJECTIVE To test whether atenolol (a long acting beta blocker) and metoprolol (a short acting beta blocker) are associated with equivalent reductions in risk for elderly patients undergoing elective surgery. DESIGN Population based, retrospective cohort analysis. SETTING Acute care hospitals in Ontario, Canada, over one decade. PARTICIPANTS Consecutive patients older than 65 who were admitted for elective surgery, without symptomatic coronary disease. MAIN OUTCOME MEASURE Death or myocardial infarction. RESULTS 37,151 patients were receiving atenolol or metoprolol before surgery, of which the most common operations were orthopaedic or abdominal procedures. As expected, the two groups were similar in demographic characteristics, medical therapy, and type of surgery. 1038 patients experienced a myocardial infarction or died, a rate that was significantly lower for patients receiving atenolol than for those receiving metoprolol (2.5% v 3.2%, P < 0.001). The decreased risk with atenolol persisted after adjustment for measured demographic, medical, and surgical factors; extended to comparisons of other long acting and short acting beta blockers; was accentuated in analyses that focused on patients with the clearest evidence of beta blocker treatment; and reflected the immediate postoperative interval. CONCLUSIONS Patients receiving metoprolol do not have as low a perioperative cardiac risk as patients receiving atenolol, in accord with possible acute withdrawal after missed doses.