The regulation in Japan takes two types of strategy in evaluating solvent vapor concentrations in solvent workplaces (SWPs); one follows the grid sampling trategy to take samples at 5 crosses in the SWP, and a geometric mean and a geometric standard deviation of the concentrations were figured out as representative parameters of solvent concentrations in the workplace. The other approach is to take one grab air sample at the site where the worker's vapor exposure concentration is empirically estimated to be the highest (the estimated highest concentration or EHC) in the SWP 1, 2) .Both of the two approaches, the grid sampling for a geometric mean and a geometric standard deviation and the grab sampling for EHC, are important because the geometric mean and the geometric standard deviation indicate the representative workplace vapor concentration (RWC) for the SWP and the sampling site-dependent variation in the concentrations (VWC) within the SWP, respectively, whereas EHC suggests the possible highest exposure condition for a worker, e.g., a foreman in the Abstract: The present study was initiated to examine the relationship between the workplace concentrations and the estimated highest concentrations in solvent workplaces (SWPs), with special references to enterprise size and types of solvent work. Results of survey conducted in 1010 SWPs in 156 enterprises were taken as a database. Workplace air was sampled at 5 crosses in each SWP following a grid sampling strategy. An additional air was grab-sampled at the site where the worker's exposure was estimated to be highest (estimated highest concentration or EHC). The samples were analyzed for 47 solvents designated by regulation, and solvent concentrations in each sample were summed up by use of additiveness formula. From the workplace concentrations at 5 points, geometric mean and geometric standard deviations were calculated as the representative workplace concentration (RWC) and the indicator of variation in workplace concentration (VWC). Comparison between RWC and EHC in the total of 1010 SWPs showed that EHC was 1.2 (in large enterprises with>300 employees) to 1.7 times [in small to medium (SM) enterprises with 300 employees] greater than RWC. When SWPs were classified into SM enterprises and large enterprises, both RWC and EHC were significantly higher in SM enterprises than in large enterprises. Further comparison by types of solvent work showed that the difference was more marked in printing, surface coating and degreasing/cleaning/wiping SWPs, whereas it was less remarkable in painting SWPs and essentially nil in testing/research laboratories. In conclusion, the present observation as discussed in reference to previous publications suggests that RWC, EHC and the ratio of EHC/WRC varies substantially among different types of solvent work as well as enterprise size, and are typically higher in printing SWPs in SM enterprises.