Hip stresses are generally believed to influence whether a hip develops osteoarthritis (OA); similarly, various osteotomies have been proposed to reduce contact stresses and the risk of OA. We asked whether elevated hip contact stress predicted osteoarthritis in initially asymptomatic human hips. We identified 58 nonoperatively treated nonsubluxated hips with developmental dysplasia (DDH) without symptoms at skeletal maturity; the control group included 48 adult hips without hip disease. The minimum followup was 20 years (mean, 29 years; range, 20-41 years). Peak contact stress was computed with the HIPSTRESS method using anteroposterior pelvic radiographs at skeletal maturity. The cumulative contact stress was determined by multiplying the peak contact stress by age at followup. We compared WOMAC scores and radiographic indices of OA. Dysplastic hips had higher mean peak contact and higher mean cumulative contact stress than normal hips. Mean WOMAC scores and percentage of asymptomatic hips in the study group (mean age 51 years) were similar to those in the control group (mean age 68 years). After adjusting for gender and age, the cumulative contact stress, Wiberg center-edge angle, body mass index, but not the peak contact stress, independently predicted the final WOMAC score in dysplastic hips but not in normal hips. Cumulative contact stress predicted early hip OA better than the Wiberg center-edge angle.