We compare categories of physical activity, based upon self-reported leisure time and occupational activity, with categories of activity based upon coded job titles using the Dictionary of Occupational Titles. For people under 65 years of age, self-reportedtotal activity, total nonintense activity, and occupational activity were not strongly associated with coded occupational activityin men (Spearman r = 0.36, 0.27, and 0.32, respectively); even weaker associations were noted in women (r = 0.23, 0.27, and0.10). Associations between total activity and coded occupational activity were generally weaker for people who were 65 or older.Despite these weak associations, all indicators of physical activity assessed were associated with a decreased risk for colon cancerin men. Comparing those in the highest category of activity with those in the lowest, the OR for total self-reported activity was0.7; for self-reported occupational activity, 0.7; and for coded occupational activity, 0.6. Whereas self-reported total activity wasassociated with colon cancer among women, estimates of activity based upon occupation (both self-reported and coded from job