This cross-sectional study was undertaken after the discovery of cobalt-related fibrosing alveolitis and bronchial asthma in diamond polishers occupationally exposed to cobalt. A total of 194 workers from 10 diamond polishing workshops and 59 workers from three other workshops from the diamond industry (control subjects) were studied; a questionnaire was administered and spirometry was performed. Cobalt exposure was assessed by environmental air sampling using both area and personal sampling and by measuring urinary cobalt concentration. When considered on a workshop basis, these environmental and biologic indices of exposure correlated well with each other. These measurements led to the definition of three cobalt exposure categories: a no exposure group, a low exposure group, and a high exposure group. The high exposure was, however, still below the present threshold limit value for cobalt (50 micrograms/m3). Spirometry showed that indices of ventilatory function (FVC and FEV1) were significantly (p less than 0.05) lower in the group with the highest exposure to cobalt. These differences were not due to differences in smoking habits. The results were confirmed when the data were analyzed by covariance analysis of lung function indices against smoking status, taking mean cobalt exposure in each workshop as a covariate, and when the data were analyzed by multiple regression analysis. This analysis showed that cobalt exposure correlated with decreased pulmonary function. Our results suggest that, during diamond polishing, exposure to cobalt at levels below the current threshold limit value is associated, on a group basis, with measurable effects on lung function parameters. The relationship between these cross-sectional epidemiologic findings and the occurrence of overt lung disease in diamond polishers remains to be established.