There is no consensus regarding treatment of periprosthetic shoulder infections. We retrospectively reviewed 17 patients diagnosed with a periprosthetic shoulder infection. Patient demographics, preoperative diagnostics, therapeutic management and functional outcome were evaluated. The Constant-Murley score (CMS), Simple Shoulder Test (SST), Visual Analogue Score (VAS) and Disabilities of the Arm, Shoulder and Hand score (DASH) were used to assess clinical outcome. Pre-and intraoperative culture results and laboratory data, including C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), were analyzed. Three patients were treated by two-stage revision arthroplasty, 5 by resection arthroplasty with implantation of a cement spacer, 8 by resection arthroplasty without spacer and one patient underwent polyethylene exchange and serial debridement. The mean follow-up was 4.7 years (range : 1-93). The CMS was 27.8 for the resection arthroplasty group, 22.7 for the two-stage revision group and 20.6 for the resection arthroplasty with spacer group. No patients received chronic antibiotic suppression. Mean CRP value was 3.7 mg/L (range: 0.2 -11.1). Infection was monobacterial in 8 patients and polymicrobial in 9. The most common organisms were Coagulase negative staphylococcus (CNS) (13/17) and Propionibacterium spp. (7/17). Complications included two humeral fractures. At a mean follow-up of 4.7 years, all but one patient were considered free of infection. Worst functional results were seen with the implantation of a definitive cement spacer. Two-stage revision arthroplasty remains the gold standard in chronic infections, but is associated with a high complication rate. One-stage revision to a reverse shoulder arthroplasty (RSA) is an attractive alternative in selected cases. A surgical treatment algorithm for infected shoulder arthroplasty is proposed.