OBJECTIVES: To determine the influence of obesity on revision rates and mortality after primary elective hip and knee arthroplasty in Germany. MATERIALS AND METHODS: In the German Arthroplasty Registry (EPRD) there were 403,073 elective total hip arthroplasties (THA), 320,913 bicondylar total knee arthroplasties (TKA) and 48,480 unicondylar knee arthroplasties (UKA) with valid BMI available for analysis. Cumulative revision rates and 1‑year mortality was calculated for BMI groups. RESULTS: There were increased revision rates with increasing BMI, most distinctive for septic revisions in THA during the first year after surgery (non-obese 0.7%, obesity grade 1 1.3%, grade 2 2.1%, grade 3 4.2%). In TKA and UKA this increase was generally less pronounced. Age- and gender-standardized mortality was lower than expected in most groups, but also increasing with increasing BMI. CONCLUSION: The risk for revisions in obese patients is more distinctive in hip arthroplasty than in knee arthroplasty and is especially high in morbidly obese patients (BMI ≥ 40 kg/m