The study evaluates the factor structure and predictive validity of the symptom questionnaire EUROQUEST (EQ) that had been developed with the goal of simplifying the evaluation of health effects associated with long-term solvent exposure. The EQ was added to the normal evaluation procedures for 118 male patients with suspected solvent-induced toxic encephalopathy (TE) referred to seven Swedish clinics of occupational medicine during an 18-month period. EQ was also completed by 239 males from a random sample of 400 Swedish males aged 25-64 years selected from the general population and a sample of 559 occupationally active male spray painters aged 25-64 years. Factor and item analyses of EQ responses were performed. Ordinary least square regression analysis was used to evaluate sensitivity and correlation to evaluate the specificity of EQ and the separate components. Questions concerning memory and concentration symptoms alone showed better sensitivity than the other five EQ dimensions singly or combined for the entire EQ and for a subset of questions approximating Q16, a widely used organic solvent symptom screening questionnaire. However, the diagnosis of TE required information in addition to exposure and responses to EQ and Q16-like questions. The results indicate that the subset of EQ questions concerning memory and concentration might replace the more cumbersome EQ and less sensitive Q16 in screening for TE, although none of the screening instruments alone replaces current clinical diagnostic procedures.