The relationship between the 8-h time-weighted average (TWA) intensity of exposure to N,N-dimethylformamide (DMF) vapor (with little possibility of skin contact with liquid DMF) and the subsequent excretion of N-monomethylformamide (MMF) precursor in shift-end urine samples was examined in 116 workers exposed to DMF and 92 workers exposed to DMF in combination with toluene. Urinary MMF level was examined also in 42 non-exposed subjects. The TWA vapor concentration in breathing zone air of each worker was successfully measured by means of a recently developed diffusive sampler in which water was used as an absorbent. The examination of gas chromatographic (GC) conditions for MMF determination showed that the formation of MMF was not saturated when the injection port temperature was set at 200 degrees C, reached a plateau at 250 degrees C, and showed no additional increase at 300 degrees C. There was a linear relationship between DMF in air and MMF in urine with a regression equation of y = 1.65 x + 1.69 (r = 0.723, P less than 0.01), where y is MMF (unit; mg/l, uncorrected for urine density) in urine and x is DMF (ppm) in air, when only those exposed to DMF were selected, and the injection port temperature was set at 250 degrees C. From this equation, it was possible to estimate that about 10% of the DMF absorbed will be excreted into urine as the MMF precursor. The slope of the regression line was significantly smaller among those exposed to DMF and toluene in combination as compared with those with DMF exposure only.