In some population-based studies, a shorter median survival was observed in peritoneal as compared with pleural, malignant mesothelioma, but in others, longer median survival times or higher proportions of long-term survivors were reported. Statistical instability could have caused these differences. We analyzed survival in peritoneal mesothelioma in a large and unselected population-based case series. Cases (338) registered from 1990 to 2001 by 9 Italian regional mesothelioma registries contributing to the network of the National Mesothelioma Registry were followed until December 31, 2005. Univariate (Kaplan-Meier) and multivariate (Cox proportional hazards regression) analyses of survival were performed according to selected individual characteristics, including limited treatment information in a subset of 194 cases. The results were compared with those obtained in a parallel study on pleural mesothelioma cases. Epithelioid histotype, younger age at diagnosis and, to a lesser degree, gender (women), and being diagnosed in a hospital with a thoracic surgery unit positively and significantly affected survival. The effect of treatment was positive but not statistically significant. No trend in the risk of death according to calendar period of diagnosis was present. Peritoneal mesothelioma cases had shorter median survival time than pleural cases, but a larger proportion of long-term survivors. Survival patterns after peritoneal and pleural mesothelioma differed markedly. Treatment was not associated with a statistically significant improvement in survival, but our study included cases first diagnosed before the introduction of the most recent therapeutic approaches. This provides a large historical comparison for future studies on survival trends at the population level.