The role of chemotherapy for unresectable malignant mesothelioma is unclear. The aims of the present study were to evaluate the methodological quality of published papers relative to chemotherapy or immunotherapy in malignant mesothelioma and to aggregate, for trials having a similar methodology, the response rates in order to identify the most active chemotherapeutic drugs and regimens. The literature relative to this topic, published between 1965 and June 2001 was reviewed. A methodological qualitative evaluation was performed according to the European Lung Cancer Working Party scale, specifically designed for phase II trials. A study was considered as potentially positive if the upper limit of the 95% confidence interval (CI) of the response rate was greater than 20% and positive if the lower limit of the 95% CI was >20%. Eighty-three studies (88 treatment arms) were eligible for the systematic review. Fifty-three arms were considered as positive or potentially positive. No statistically significant difference in the methodological quality was observed between negative and positive studies. Studies were aggregated in four groups according to the presence of cisplatin and/or doxorubicin in the treatment regimen. The combination of cisplatin and doxorubicin had the highest response rate (28.5%; P<0.001). Cisplatin was the most active single-agent regimen. Our systematic qualitative and quantitative overview of the literature suggests that the most active chemotherapeutic regimen, in term of objective response rate, is the combination of cisplatin and doxorubicin and the best single-agent is cisplatin. The combination of these two drugs can be recommended as control arm for future randomised phase III trials.