The significance of 3 commonly used nonspecific exposure descriptors was investigated. Subjects with early COPD, employed full or part-time at baseline (5213) in the multisite Lung Health Study, participated. At the baseline and annually for 5 subsequent years, questionnaires asked about current employment status & three exposure measures: 1. Exposure to dust (D); 2. exposure to fumes (F); 3. "use of a mask" (M). In addition, each subject listed occupation, job title, and industry using free text. For a stratified random sample, the free text descriptions were rated by experts for 1. Rated Dust exposure; 2. rated odor exposure; 3. rated mask need; 4. rated combustion products exposure. Overall, dust exposure (d), fume exposure (F), and mask (M) use were reported by 20, 18, 5% respectively. Corresponding expert ratings were higher for subjects reporting D, F, M. There was a high degree of concordance between reported D and F. Among mask users, fewer women reported D or F than did men. Consistency of report on repeat questionnaire was related to extent of exposure as reflected by expert ratings. Expert odor ratings were significantly higher among those reporting F; in subjects discordant for D and F, odor ratings were higher in those reporting F. 1. The 3 commonly used questions are sueful and relate to exposure situations, 2. The term "fume" is ambiguous, and may often be interpreted to mean "odor," 3. "Fume" is partially redundant to "Dust". 4. The significance of "Mask" differs between men and women. 5. Where possible, these questions should be repeated on several occasions.