Sixty workers, consecutively admitted due to suspicion of solvent-induced chronic toxic encephalopathy (CTE), were investigated with pure-tone audiometry, determination of speech recognition of monosyllabic words and distorted speech and cortical response audiometry (CRA). Eighteen workers not exposed to occupational solvents and noise were also investigated. The scores in the distorted speech test were significantly lower and the CRA latencies were significantly longer in the solvent group than in the control group. There was no difference between the groups in the pure-tone and monosyllabic speech recognition tests. In the solvent group, 19 subjects had one or several pathological audiological test results (values exceeding the mean result of the control group by 2 SD). Independently of the audiological examination all the workers in the solvent group underwent the traditional clinical assessment of CTE, which is based on symptoms, history of exposure, clinical neurological examination and a neuropsychological investigation. They were classified in three groups--CTE, incipient CTE and non-CTE. There was no correlation between these groups and the audiological test results. A previous report on vestibular pathology in the same group of subjects and the present investigation on hearing deficits suggest that long-term exposure to solvents causes disturbances of the central pathways in the otovestibular system. Hitherto, no attention has been paid to these disturbances in the definition of the CTE syndrome.