In order to determine the possible actual exposure to benzene in the shoe industry from industrially used solvents, glues, and paints containing benzene as a nondeclared constituent, phenol in urine and benzene in blood, as indices of internal exposure to benzene, were measured in workers (N = 33). Since toluene, in contrast to benzene, is declared as a constituent in several glues, toluene in the blood of workers was also analysed. All analyses were performed using gas chromatography. Urine samples were collected on Monday morning (MI) before work and on Wednesday (WI) before and (WII) after work. Venous blood samples were taken on Wednesday only, 1/2 hour after work. There was no difference in the phenol concentrations of MI and WI, while the phenol concentration of WII was about twice as high as that in WI. In all blood samples, benzene was found, as well as toluene, which was about four times higher in comparison with benzene. A correlation (r = 0.465; p less than .01) was found between the difference in pre- and postshift phenol concentrations (WII-WI) in urine and the benzene concentrations in blood. The results presented show that a trace amount of benzene, which is often not declared as a constitutent in industrially used chemicals, could be a source of marked exposure to benzene. It can also be concluded that changes in phenol in urine (if preshift and postshift samples are taken) might be a sufficiently sensitive parameter to assess exposure to benzene even when other data concerning the presence of benzene in the working atmosphere are not available.