Latency periods (time intervals elapsing between first exposure to asbestos and death) were examined in 421 cases of malignant pleural mesothelioma, diagnosed in the Trieste-Monfalcone area, Italy. Occupational data were collected from the patients or from their relatives by personal or telephone interviews. Routine lung sections were examined for asbestos bodies in 370 cases. Latency periods, calculated in 312 cases, ranged from 14 to 72 years (mean 48.7, median 51). Latency periods differed significantly from one occupational group to another. Mean latency periods were 29.6 among insulators, 35.4 among dock workers, 43.7 in a heterogeneous group defined as various, 46.4 in non-shipbuilding industry workers, 49.4 in shipyard workers, 51.7 among women with a history of domestic exposure to asbestos, and 56.2 in people employed in maritime trades. The ANOVA test indicated a correlation between latency periods and occupational groups. Latency periods in people with asbestos bodies visible in routine lung sections did not differ from those seen in cases with no evidence of asbestos bodies. These data suggest that intensity of exposure is a relevant, but not the only, factor in determining the duration of latency periods.