The use of isolated perfusion with melphalan in an adjuvant setting for stage I extremity melanoma continues to be controversial. In this article, 2 retrospective studies compare patients undergoing perfusion from Groningen, the Netherlands, with matched controls from the Sydney Melanoma Unit in Australia and from 5 hospitals in the Netherlands and Westphalia (a region of Germany bordering the Netherlands). All patients underwent wide local excision for a primary extremity melanoma of 1.5 mm or more in thickness. In the first study, women with a melanoma of the leg (excluding the foot) and treated with perfusion had a significantly higher survival than women treated with local excision alone. In the second study, a proportional hazards regression analysis for recurrent disease and survival identified no statistically significant difference in favor of adjuvant perfusion. Two other studies from the literature evaluating the efficacy of perfusion are examined. Patients treated at the M.D. Anderson Cancer Center, Houston, with wide local excision and perfusion are compared retrospectively with a group of patients individually matched for prognostic factors from the University of Alabama in Birmingham, U.S.A. and the Sydney Melanoma Unit. A small subset of patients with tumors greater than 2 mm thick benefited from perfusion. A prospective randomized study from the University Hospital, Cologne, West Germany found benefits for all perfused patients with tumors greater than 1.5 mm but their patients did not appear to be individually matched for tumor thickness. It is concluded after critical appraisal of all these studies that there is insufficient convincing evidence to date of an advantage of perfusion over wide local excision alone.