Background Previous observational studies that have examined the association of allopurinol with myocardial infarction (MI) have provided contradictory results. One study showed that allopurinol reduced the risk, while another study showed an increased risk with allopurinol. Therefore, our objective was to assess whether allopurinol use is associated with a reduction in the risk of MI in the elderly. Method We used the 2006-2012 5 % random sample of Medicare beneficiaries to study the association of new allopurinol initiation and the risk of incident MI in a cohort study. Multivariable-adjusted Cox regression models adjusted for age, gender, race, and Charlson index, in addition to various cardio-protective medications (beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, diuretics, statins). We calculated hazard ratios (HRs) with 95 % confidence intervals (CIs). Sensitivity analyses adjusted for coronary artery disease (CAD) risk factors, including hypertension, hyperlipidemia, diabetes, and smoking. Results Of the 29,298 episodes of incident allopurinol use, 1544 were associated with incident MI (5.3 % episodes). Allopurinol use was associated with reduced hazards of MI, with a HR of 0.85 (95 % CI, 0.77 to 0.95). Compared to no allopurinol use, longer durations of allopurinol use were associated with a lower HR of MI: 1-180 days, 0.98 (95 % CI, 0.84 to 1.14); 181 days to 2 years, 0.83 (95 % CI, 0.72 to 0.95); and >2 years, 0.70 (95 % CI, 0.56 to 0.88). Other factors associated with a higher hazard of MI were: age 75 to <85 years and ≥85 years, male gender, higher Charlson index score, and the use of an ACE inhibitor. Adjustment for CAD risk factors confirmed these findings. Conclusion Incident allopurinol use was associated with a reduction in the risk of incident MI in the elderly. Longer durations of allopurinol use reduced the risk of incident MI incrementally. Future studies should assess the underlying mechanisms for MI prevention and assess the risk-benefit ratio for allopurinol use.