A quantitative exposure-estimating algorithm for benzene and total hydrocarbons was developed for a case control study of petroleum marketing and distribution workers. The algorithm used a multiplicative model to adjust recently measured quantitative exposure data to past scenarios for which representative exposure measurement data did not exist. This was accomplished through the development of exposure modifiers to account for differences in the workplace, the materials handled, the environmental conditions, and the tasks performed. Values for exposure modifiers were obtained empirically and through physical/chemical relationships. Dates for changes that altered exposure potential were obtained from archive records, retired employee interviews, and from current operations personnel. Exposure modifiers were used multiplicatively, adjusting available measured data to represent the relevant exposure scenario and time period. Changes in exposure modifiers translated to step changes in exposure estimates. Though limited by availability of data, a validation exercise suggested that the algorithm provided accurate exposure estimates for benzene (compared with measured data in industrial hygiene survey reports); the estimates generally differed by an average of less than 20% from the measured values. This approach is proposed to quantify exposures retrospectively where there are sufficient data to develop reliable current era estimates and where a historical accounting of key exposure modifiers can be developed, but where there are insufficient historic exposure measurements to directly assess historic exposures.