A prospective investigation is described which sought to test the role of illness cognitions in determining patient responses to a surgical intervention for osteoarthritis. Illness cognitions were assessed amongst a consecutive sample of patients with osteoarthritis of the knee or hip prior to undergoing joint replacement surgery. Functional activity and depression were assessed pre-operatively, and at 3 and 9 months post surgery. At pre-operative assessment, functional activity and depression were univariately associated with the perceived consequences of osteoarthritis. Path analyses using longitudinal data demonstrated that illness cognitions had predictive value in explaining outcomes. Depression at 3 months was associated with higher pre-operative perceived control beliefs suggesting that patients who have high control pre-operatively may be at risk for temporary depressed mood in the immediate aftermath of surgery. This effect was not maintained at 9 months. Depression at 9 months was lower amongst patients who were more active at 3 months, who did not attribute their condition to wear and tear and who had higher expectations of surgery. Functional activity at 9 months was higher amongst those who did not attribute their condition to growing older and who perceived more control over symptoms. Socio-demographic variables were not associated with change in functional activity or depression over the course of surgery. The results provide support for an illness cognition approach in explaining functional activity and depression outcomes following surgery.