Résumé : Objective: To investigate the association between abacavir (ABC) use and recurrent myocardial infarction (MI) among HIV-positive people with a prior MI. Design: International multicohort collaboration with follow-up from 1999 to 2016. Methods: The rate of recurrent MI was described among D:A:D participants who experienced an index MI whilst in the study, and who remained under follow-up beyond 28 days after this MI. Follow-up was considered to the date of next MI, death, 1 February 2016 or 6 months after last clinic visit. Poisson regression models considered associations between recurrent MI and exposure to ABC (use at index MI, current post-MI exposure and cumulative exposure), before and after adjusting for calendar year. Results: The 984 individuals who experienced an index MI during the study (91.3% male, median age 51 at index MI) were followed for 5312 person-years, over which time there were 136 recurrent MIs (rate 2.56/100 person-years, 95% confidence interval 2.13-2.99). Rates were 2.40 (1.71-3.09) and 2.65 (2.10-3.21)/100 person-years in those who were and were not on ABC, respectively, at the index MI, and 2.90 (2.01-3.78) and 2.44 (1.95-2.93)/100 person-years in those who were and were not currently receiving ABC, respectively, post-MI. No association was seen with recurrent MI and either cumulative exposure to ABC [relative rate 0.86 (0.68-1.10)/5 years], receipt of ABC at index MI [0.90 (0.63-1.29)] nor recent post-MI exposure to ABC [1.19 (0.82-1.71)]. Conclusion: Among people with a previous MI, there was no evidence for an association between use of ABC post-MI and an elevated risk of a recurrent MI.