A case-control study was conducted to evaluate the risk of chronic beryllium disease (CBD) and beryllium sensitization (SENS) associated with various levels of historical beryllium exposure at the Rocky Flats nuclear weapons facility. Fifty CBD and 74 SENS cases were matched to controls of the same age group, race, gender, and smoking status. A job exposure matrix was developed from job history data and fixed airhead (FAH) exposure data available from 1960 to 1988. Job titles and building areas were assigned factors based on exposure relative to a machinist in the Building 444 Beryllium Shop. Concurrence on these factors was obtained from past and present Rocky Flats industrial hygienists. Using the matrix, long-term mean and cumulative exposures were estimated for each subject. Both exposure estimates (p < 0.0001) and years of employment (p = 0.010) were found to be significantly higher for CBD cases as compared with their controls, but not so for the SENS cases as compared with their controls. Logistic regression analyses showed statistically significant relationships between both cumulative and mean exposure and CBD, but not for SENS. These findings suggest that reduced worker exposures might lower the future incidence of CBD, but may not necessarily lower the incidence of SENS.