We conducted a proportional mortality study of 1043 deaths among men who took part in an antimalarial campaign in Sardinia, Italy from 1946 to 1950. DDT comprised 94% of the insecticide used during the campaign, and was sprayed over the soil of the entire region at an average concentration of 10 mg/m2, as well as in all dwellings and animal shelters. Expected deaths were derived from the proportional mortality rates of the general Italian male population, specific by cause, 5-year age groups, and 5-year calendar periods in the period from 1956 to 1992. The proportional mortality ratio (PMR) for cardiovascular diseases was significantly decreased, while nonmalignant respiratory diseases showed a 22% increase in risk of borderline statistical significance. Significant increases in risk among workers exposed to DDT in application or inspection jobs were observed for liver and biliary tract cancer (PMR = 228; 95% C.I. = 143-345) and multiple myeloma (PMR = 341; 95% C.I. = 110-795). The PMR for myeloid leukemia was also increased (PMR = 189; 95% C.I. = 38-552), although it was not statistically significant. PMRs for liver and biliary tract cancer and myeloid leukemia were also elevated among workers who did not have direct occupational contact with DDT (liver and biliary cancer: PMR = 210; 95% C.I. = 117-346; myeloid leukemia: PMR = 170; 95% C.I. = 19-614). No trends occurred according to length of employment in exposed jobs. These preliminary results are somewhat in agreement with experimental studies in rodents and previous epidemiologic findings. Expansion of the cohort to include all applications, and collection of information to improve exposure assessment is needed to clarify these findings.