The validity and reliability of occupational physical activity data from survey instruments were determined in 75 men and women (age 23 to 59 years) in white-collar jobs. Data were validated against measures of cardiorespiratory fitness, body fatness, motion detection, pulmonary function, and 12 days of occupational physical activity records. More than 90% of occupational physical activity was spent in light-intensity sitting, standing, and walking. Test-retest reliability was high for most occupational questions (r > or = .63, P < .05). Correlation coefficients between questionnaire data and validation criteria that reflect heavy-intensity physical activity were low (r < .17, P > .05). The modified Tecumseh Occupational Questionnaire and Seven-Day Recall, which classified occupational physical activity in hours per week worked and average level of ratio of associated metabolic rate for a specific activity divided by the resting metabolic rate, on the job, correlated best with physical activity records (r = .11 to .47). Validation of these self-administered questionnaires in populations with a greater diversity of jobs and occupational energy requirements is needed.