OBJECTIVE: This study aimed to obtain an upper estimate of any survival benefit conferred by resection in patients with a diagnosis of malignant pleural mesothelioma. METHODS: We analysed published data concerning survival from diagnosis among four groups of patients with mesothelioma, identified by ascending level of intervention: (A) no surgery; (B) thoracotomy but no resection; (C) resection but no adjuvant treatment; and (D) resection as part of multimodality treatment. Mean survival was estimated for each of these four groups. Mean survival was also estimated for all those having resection (groups C and D) and for all those not having a resection (groups A and B). RESULTS: Mean survival was 16.8, 17.8 and 17 months for those having no surgery, thoracotomy alone and resection with no adjuvant treatment respectively (groups A, B and C) and 32.9 months for those having multimodality treatment (group D). Mean survival was 25.6 months in those who had resection and 17.1 months in those that did not. The survival advantage of any management that included surgical resection was estimated as being no more than 9 months. This is the most optimistic estimate and requires all observed differences in survival to be attributed to the effect of treatment and none to selection for treatment. Furthermore, within this upper estimate is included any benefit from other components of multimodality treatment. CONCLUSIONS: Given the burden of morbidity of resection in the management of pleural mesothelioma, this most optimistic estimate of the magnitude of any survival benefit should be taken into account in any policy decision, in clinical trial proposals and in strategies adopted by clinical teams.