A case-control study of chronic neurological and psychiatric disease and occupational exposure to solvents was carried out in eight automobile assembly plants. Cases included 299 subjects who were granted medical disability retirement in 1980-8. Two control groups were selected, the first from those granted retirement in the same period because of medical disability from causes unrelated to solvent exposure. The second included hourly employees from the plant population. In these facilities, solvent exposures tended to be short term and low level, although common: the average duration of exposure was 2.3 years; about 41% experienced at least one day of exposure. Of those exposed, 46% had less than one year of exposure. Results for all psychiatric disease combined (273 cases) suggested that cases had lower exposures than either control group, regardless of how exposure was expressed. Results could not be explained by conventional confounding exposures or characteristics or by usual manifestations of the healthy worker effect. By contrast, chronic neurological disease, and multiple sclerosis in particular, seemed to be associated with exposure, although few cases were identified and observed increases in risk were not statistically significant.