A survey of solvent was conducted for 196 unit work areas in 95 plants in 1994 to 1996 in Hiroshima Prefecture, Japan. The survey had been repeated every 6 months (i.e., twice a year) during the 3-year period. Sampling and analysis of the solvent vapors were carried out after national protocols set by the regulation. Toluene was most frequently detected regardless of the type of solvent work (except for degreasing), whereas the second- and the third-most common solvents varied depending on the type of solvent works. Among chlorinated hydrocarbon solvents for degreasing, dichloromethane was most widely used. Solvent concentrations were generally low as none of the median concentrations exceeded corresponding Administrative Control Levels set by the regulation, either individually or even when the assumption of additiveness was applied. Among the 1176 cases analyzed, 80% of the unit work areas were evaluated as adequate (i.e., classified as Class I). Furthermore, about 57% stayed in Class I throughout the 3 years, suggesting that solvent exposure conditions were generally quite stable. In regulatory evaluation by classification, A-sampling was decisive in most cases, whereas the role of B-sampling was limited.