In view of the historic importance of smelter workers in the field of occupational medicine, it is surprising that until very recently little data was available on the mortality experience of these workers. The problem in most studies lies in identifying the smelter workers, because smelting, strictly speaking, refers to the melting of ores for the purposes of recovering metals, whereas smelters sometimes perform the operations of roasting, calcining, sintering, converting, and refining. These distinctions are not made in most mortality studies. Most mortality studies of smelter workers conducted to date have shown some excess in lung cancer. For lead, copper, cadmium, and nickel smelters a different etiologic agent has been proposed for each. These different explanations arise partly from different initial perspectives in conducting the studies. In this paper, data are presented on a current historical-prospective study of males who worked a year or more during the period January 1, 1940 to December 31, 1964 at a copper smelter in Tacoma, Washington. This smelter (and refinery) handled a copper ore with a relatively high arsenic content and produced arsenic trioxide as a by-product. Overall 97.2% of the original study population was traced through 1976. Of the 1,061 who were found to have died, death certificates were obtained for 1,018, or 96%. For all causes of death, the mortality rates in this cohort, expressed as a Standardized Mortality Ratio (SMR), were 3.5% higher than that expected based on the United States white male mortality experience. A total of 104 respiratory system cancers were observed compared to 54.6 expected (SMR = 190.5, p less than .05). Respiratory cancer rates were found to be elevated in both smokers and nonsmokers. Overall, a gradual rise in SMR's for respiratory cancer was observed with increasing duration of exposure but not with an increasing interval from onset of exposure. This observation is consistent with the notion that the effects of arsenic on cancer incidence disappear with time. This phenomenon has also been observed for cigarette smokers and chromate workers. Additional analyses are planned for the Tacoma cohort, which will examine mortality according to process as well as by exposure to arsenic and other contaminants. The importance of studies of workers at other kinds of smelters and refineries is stressed.