Cause-specific mortality patterns among Rhode Island jewelry manufacturing workers, as identified on death certificates from 1968 to 1978, were examined using the proportionate mortality ratio (PMR) method. Among males, elevated PMRs were observed for nonmalignant kidney disease (PMR = 163; Observed deaths [Obs] = 19; p less than or equal to .05), liver cancer (PMR = 297; Obs = 6; p less than or equal to .05), drug dependence (PMR = 379; Obs = 5; p less than or equal to .05), and accidental poisonings (PMR = 274; Obs = 8; p less than or equal to .05). All but one of the deaths from accidental poisonings involved drugs or drugs and alcohol. Among females, elevated PMRs were observed for stomach cancer (PMR = 174; Obs = 20; p less than or equal to .01), peptic ulcer (PMR = 235; Obs = 8; p less than or equal to .05), diseases of the skin and subcutaneous tissue (PMR = 383; Obs = 5; p less than or equal to .05), and drug dependence (PMR = 674; Obs = 3; p less than or equal to .05). Exposure to known renal toxins (heavy metals and solvents) used in the jewelry industry may account for the excess deaths from kidney disease. The elevated PMR for liver cancer may be due to exposure to solvents (trichloroethylene, perchloroethylene, carbon tetrachloride) that cause liver cancer in animals. The elevated PMRs for drug dependence and accidental poisonings may be due to the socioeconomic status of jewelry workers, or to the interaction between solvents used in jewelry manufacturing and drugs, or drugs and alcohol. Because of the lack of information about the specific occupational exposures of the decedents, this should be viewed as an exploratory investigation requiring further follow-up.