A cross-sectional study was performed to evaluate the effects of chronic low-dose solvent on neurobehavioral performance of 118 male car painters. A control group of 113 workers matched for age was selected from different sections of the factory. The mean age and the mean duration of employment were 33 years and 6.7 years in both groups. Mean years of education were 11.4 years in car painters and 11.8 years in controls. Each worker completed a medical and occupational questionnaire and four tests of Swedish performance evaluation system. These included simple reaction time, symbol digit, digit span and finger tapping speed. Althougth the mean duration of employment was 6.7 years, comparison of mean performance showed a significantly poorer performance on simple reaction time (p<0.05), symbol digit(p<0.01) and digit span(p<0.05) in car painters. In univariate analysis, age and educational level contributed to poorer performance on symbol digit and digit span. Smoking appeared to slow finger tapping speed in car painters. Performance of four tests of car painters exposed to high level of solvent was poorer than that of car painters exposed to low level. In multiple regression models, controlling for age, alcohol, smoking and shift work, solvent exposure was found to be associated with performance of simple reaction time, symbol digit and digit span and exposure to high level of solvent was related to poorer performance of symbol digit and digit span.