OBJECTIVES: The aim of this study was to determine the dose-effect relationship between solvent exposure and acute neurobehavioural effects at the worksite. METHODS: In a balanced design, ten workers in a Swiss foundry were monitored for 15 days at ten different times during work. Urine samples were taken in the morning and at the time of examination, and personal exposure to isopropanol and methylformate was measured with active samplers. Neurobehavioural tests such as postural balance (bipedal, bipedal blind, monopedal), simple reaction time and digit span of the Neurobehavioural Evaluation System (NES2) and a combined memory and reaction-time test, the combi-test, were performed. A rating of well-being, and the last consumption of alcohol, caffeine, nicotine and medication were reported. RESULTS: Average environmental concentrations of isopropanol were at 44 ppm ( +/- 16 ppm), and at 36 ppm (+/-21 ppm) for methylformate. Maximum values of personal exposure to isopropanol reached barely the maximal allowable concentration (MAC) value (400 ppm); the methylformate personal exposure of three workers exceeded the MAC value (100 ppm). Urine concentrations of methanol were high (3.1 +/- 2.3 mg/l in the morning, 7.8 +/- 4.9 mg/l after exposure) compared with the results of other studies; concentrations of isopropanol were rather low (0.88 +/- 0.73 mg/l after exposure). CONCLUSIONS: Nevertheless, between personal exposure and biomonitoring, linear correlation was found. Methylformate exposure correlated with methanol and formic acid concentration in the urine, and isopropanol exposure with its concentration in the urine. With the neurobehavioural tests used, no solvent effect in relation to the dose could be determined.