Objective To estimate in patients with Staphylococcus aureus prosthetic joint infection after total hip arthroplasty (THA) or total knee arthroplasty (TKA) the microorganism-specific cumulative probability of treatment failure after prosthesis removal and delayed reimplantation arthroplasty. Patients and Methods All patients with S aureus THA or TKA infection, according to a strict case definition, who were treated with prosthesis removal and delayed reimplantation arthroplasty at Mayo Clinic Rochester between 1980 and 1991 were identified. The study group comprised patients who were free of infection at the time of reimplantation arthroplasty. This cohort was followed up until treatment failure, infection with another organism, prosthesis removal, death, or loss to follow-up occurred. The Kaplan-Meier survival method was used to estimate the cumulative probability of treatment failure. Results Among 120 S aureus prosthetic joint infections treated with prosthesis removal during the study period, 38 episodes (22 THA, 16 TKA) in 36 patients met the study inclusion criteria. After a median of 7.4 years (range, 0.9 year-16.4 years) of follow-up, treatment failure occurred in 1 (2.6 %) of 38 episodes 1.4 years after reimplantation arthroplasty. The 5-year cumulative probability of treatment failure was 2.8% (95% confidence interval, 0%-8.2%). Conclusions These data suggest that prosthesis removal and delayed reimplantation arthroplasty is an effective treatment to limit the recurrence of S aureus prosthetic joint infection, provided there is no evidence of infection at the time of reimplantation arthroplasty.