A significant increase in lung cancer was observed in a previous study on the mortality experience of a cohort of 1332 male workers employed between 1959 and 1980 in a resin manufacturing plant. Due to the limited exposure and an inadequate follow-up, it was not possible to make a thorough analysis of the potential association of this elevated risk with exposure to formaldehyde. The study was therefore continued and extended for a further six years (1980-1986), in order to overcome the limitations. Despite these attempts, however, there were still 219 workers whose specific exposure could not be identified. Lung cancer risk in the whole cohort (27,202 person-years) was equal to that of the local population (observed = 24; expected = 23.9). Among those definitely exposed to formaldehyde, 6 lung cancer cases were observed and 8.7 were expected, while those with non-specified exposure exhibited an increase risk (observed = 9; SMR = 211); they were mainly short-term workers employed at the beginning of operations. The previously suggested increase in haematologic neoplasms was confirmed (observed = 7; SMR = 143); the risk was highest among formaldehyde-exposed workers (observed = 3; SMR = 173). Five deaths due to primary liver cancer were observed, while 2.0 would have been expected from the local population rates (SMR = 244); the increased risk was fairly evenly distributed across the exposure categories (exposed to formaldehyde, SMR = 244; non-exposed to formaldehyde, SMR = 227; non-specified exposure = 287); however, all cases were first exposed at the age of 45 years or older. A noteworthy finding was a 50% increase in mortality from respiratory diseases. The increase was mainly apparent among those with longest and earliest exposure, employed in operations classified as involving exposures other than formaldehyde (observed = 9; SMR = 224). Overall, the results of this extended study do not provide sufficient grounds for associating work in formaldehyde resin production in this plant with increased carcinogenic risk; however, limitations in the individual exposure classification and suggestions of an increased risk for certain tumours preclude considering the study as negative. The numerous airborne irritative agents present in the plant environment appeared to have increased the risk of respiratory disease.(ABSTRACT TRUNCATED AT 250 WORDS)