Background: Intense exercise has been reported as one risk factor for hip and knee osteoarthritis (OA). Purpose: This study aimed to evaluate (i) whether former impact and non-impact athletes have more OA than expected, (ii) if former athletes have more hip or knee arthroplasties due to OA than expected, and (iii) if knee OA is driven by previous knee injuries. Methods: The prevalence of OA and arthroplasty in the hip and knee were registered in 709 former male elite athletes with a median age of 70 years (range 50-93), retired from sports for a median 35 years (range 1-63), and compared with 1368 controls aged a median 70 years (range 51-93). Both cohorts were normally distributed (Shapiro Wilk’s test). Data are presented as means with 95% confidence intervals (95% CI). Group differences in anthropometrics and lifetime factors were evaluated by Student’s t-test between means and chi-square test. Age-adjusted odds ratios (OR) was estimated by logistic regression in different models adjusted for combinations of age, body mass index (BMI; kg/m2), occupational load, and previous soft tissue knee injury. Results: The age adjusted risk of hip or knee OA was higher in former athletes (OR 1.9, 95% CI 1.5, 2.3), as was arthroplasty based on OA in either of these joints (OR 2.2, 95% CI 1.6, 3.1) (Table 1). The risk of hip OA was doubled (OR 2.0, 95% CI 1.5, 2.8) and hip arthroplasty 2.5 times higher (OR 2.5, 95% CI 1.6, 3.7) in former athletes than in controls, predominantly driven by a higher risk in former impact athletes. Also the risk of knee OA was higher (OR 1.6, 95% CI 1.3, 2.1), as was knee arthroplasty (OR 1.6, 95% CI 0.9, 2.7), driven by a higher risk in both former impact and non-impact athletes. The differences between the groups remained after adjustments for age, BMI, occupational load, and soft tissue knee injury, except for knee OA where impact sportsmen no longer were at risk (OR 1.19, 95% CI 0.83, 1.71) when adjusted for previous knee soft tissue injury while the risk remained after adjustment in non-impact athletes (OR 3.19, 95% CI 1.47, 6.91). Conclusions: Hip and knee OA and hip and knee arthroplasty are more commonly found in former male elite athletes than expected. A previous knee injury is associated with knee OA in former impact sportsmen, but not in non-impact athletes.