A technique for estimating historical occupational exposures to refractory ceramic fibers (RCFs) during manufacturing and related operations was described. The technique involved constructing an algorithm that would enable all RCF exposures occurring at the facilities through 1991 to be reconstructed. Data needed for the algorithm were collected through structured interviews with workers at the facilities, reviews of engineering data related to production processes and environmental controls contained within company records and files, reviews of historical (pre 1987) industrial hygiene dust monitoring data identified in factory records, and reviews of industrial hygiene monitoring data obtained at the factories after 1987. The dust zone (DZ) concept which involved grouping together similar work tasks, materials, processes elements, control technologies, and worker groups was applied when analyzing the industrial hygiene monitoring data. The algorithm when constructed incorporated terms representing the least squares mean of the log transformed daily time weighted average (TWA) RCF exposures, a function related to the squared standard deviation of the TWA mean exposures, the dates of implementation of engineering or process modifications that might be associated with changes in airborne RCF concentrations, and the DZs where the samples were collected. Differences in RCF exposures during various time periods in various DZs were tested for statistical significance by analysis of variance. The algorithm was able to reconstruct exposure estimates for 81 job titles for various time periods between 1953 and 1991. Estimates based on the algorithm indicated that RCF exposures at the facilities decreased over time. The maximum RCF exposure was around 10 fibers per cubic centimeter (f/cc) and occurred in carding operations during textile production in the 1950s. Engineering methods applied at later times were able to reduce this exposure to below 1f/cc. Most RF exposures measured after 1974 have not exceeded 0.6f/cc.