Exposure to toluene in two publication rotogravure plants was investigated to examine how accurately long-term exposure can be estimated on the basis of production rate. Toluene consumption was used as the measure of production rate. Continuous area monitoring was used to find a correlation between production rate and airborne level of toluene. Workers' exposure levels were first estimated by combining data on toluene concentrations in various monitoring sites with data supplied by the workers on the time spent in these areas. These calculated exposure levels were found to correlate well with the actual exposure levels obtained by breathing zone sampling. There was also a fairly high correlation between the concentration of toluene in front of the press and the consumption of toluene if the process conditions remained stable. It was, however, necessary to investigate this association separately for the situations where the degree of enclosure of the press or number of emission sources were unusual or when the workers stayed in the control rooms, which were separated from the other pressroom areas. A reasonably high correlation between the variables of the main interest, that is, the calculated toluene exposures and the consumption of toluene, was found in one of the plants investigated, whereas this correlation was low in the other plant. Even though this kind of estimation procedure does not always lead to accurate exposure levels, it helps in understanding how those are affected by the process parameters.