The investigation describes mortality of vinyl chloride exposed workers in the Montedison-Enichem plant located in Porto Marghera, near Venice, Italy. A total of 1658 workers employed from start of production (1950), present in 1956 or successively hired until 1985, were followed up between 01.01.1973 and 31.07.1999, for a total of 41.037 person years at risk: 248 deaths were observed. Mortality from all causes compared with regional population was lower than expected, (SMR 0.75; 90% CI 0.68-0.83) and from all malignant neoplasms similar to expected (SMR 0.94; 90% CI 0.81-1.09). SMR for primary liver cancer was significantly increased (SMR 2.78 90% CI 1.86-4.14). In the first year since leaving employment observed mortality was significantly above the null value for all causes (SMR 2.76; 90% CI 1.94-3.91), all malignant neoplasms (SMR 1.89; 90% CI 0.97-3.92) and cardiovascular diseases (SMR 2.37; 90% CI 1.13-4.95). Mortality rates for liver angiosarcoma (6 cases) increased with latency (trend test x 2 (1df) = 25.20 p < 0.001) and cumulative exposure (trend test x 2 (1df) = 61.00 p < 0.001), there were no cases with duration of employment less than 12 years, latency less than 10 years and for cumulative exposure less than 2.379 ppm-years. Mortality rates for hepatocellular carcinoma (12 cases) and liver cirrhosis (20 cases) showed a similar pattern for cumulative exposure. Observed mortality from lung cancer was higher than expected among those workers whose only job title was bagger. In the analysis accounting for latency, age and calendar period the RR for only bagger was 2.31 (90% CI 1.15-4.61). Mortality pattern for all causes, all malignancies and cardiovascular disease increased by time since employment, as expected in presence of a particularly strong Healthy Worker Effect (HWE). These results and the increased SMR values during the first year since leaving employment indicate that workers were selected into employment on the basis of good health conditions and early selective removal of weaker ones followed. The study results confirm the causal relationship between VCM exposure and liver angiosarcoma, and add supplementary evidence in favour of a causal explanation of the excess risk for hepatocellular carcinoma and liver cirrhosis as well as lung cancer among only baggers.