The aim of this study was to evaluate retrospectively the rate of disease progression in the asymptomatic hip of patients who had non-traumatic osteonecrosis and pain in the other hip. Of seventy-five consecutive patients with non-traumatic osteonecrosis who were initially evaluated between 1978 and 1987, forty-six (61 per cent) initially had pain in only one hip. Radiographs were made for all patients, but routine magnetic resonance images and computed tomography images were not. Twenty-two of the asymptomatic hips had radiographic evidence of involvement, but three of them were excluded because they were treated with elective prophylactic core-drilling; this left nineteen hips with radiographic evidence of involvement in the final study group. The other twenty-four asymptomatic hips had normal findings on the initial radiographs; one of these was treated with elective core-drilling, leaving twenty-three hips with normal radiographs in the final study group. All but one patient, who died at fifty-one months, were followed for at least five years or until symptoms developed in the asymptomatic hip.Of the nineteen untreated asymptomatic hips with initial radiographic evidence of involvement, five were still asymptomatic at the most recent follow-up examination. The other fourteen hips had become painful: nine, within five years after presentation, and five, more than five years after presentation. Pain or radiographic changes developed only rarely in the twenty-three asymptomatic hips with normal findings on the initial radiographs, and only one hip had both pain and radiographic changes within five years after presentation. Nineteen (83 per cent) were still asymptomatic at the most recent follow-up examination.The prolonged pain-free interval for many of the asymptomatic hips with radiographic evidence of involvement contrasts with the over-all rapid progression of disease reported for most hips with non-traumatic osteonecrosis. The present study draws attention to the fact that an adequate period of follow-up is needed for asymptomatic hips that are treated operatively in order to determine whether such intervention alters the natural history of the disease. Our results also showed that few asymptomatic hips with normal findings on initial radiographs are at risk for pain or radiographic abnormalities; when disease does develop, deterioration is slow and operative intervention is rarely indicated.