Differences in survival outcome for patients with ovarian cancer in Scotland led to an investigation of whether these differences were due to variation in presenting prognostic features or to the organisation and delivery of cancer services. A retrospective study of all 533 cases of ovarian cancer registered in Scotland in 1987 was carried out. After adjustment for age, stage, pathology, degree of differentiation and presence of ascites, survival improved when patients (1) were first seen by a gynaecologist (P < 0.05); (2) were operated on by a gynaecologist (P < 0.05); (3) had residual disease of less than 2 cm post-operatively (P < 0.001); (4) were prescribed platinum chemotherapy (P < 0.05); and (5) were referred to a joint clinic (P < 0.001). When gynaecologists operated the likelihood of smaller residual disease increased (P < 0.001). The improved survival from management by a multidisciplinary team at a joint clinic was not solely due to the prescription of platinum chemotherapy. The results of this study support the contents of the 1991 Department of Health report on present acceptable practice in the management of ovarian cancer, circulated to gynaecologists and surgeons in Scotland in 1992. The new finding that in a common cancer management by a multidisciplinary team at a joint clinic directly affects survival requires urgent attention.