Aim of the present study was to design a questionnaire to assess health related physical activity, to validate the instrument and to apply it to a population sample. Reliability of the questionnaire was evaluated by test-retest investigations with intervals of two weeks and six months. High correlations between the repeated administrations reflect a good reliability of our instrument. Only gardening and cycling, as well as the depending basic and total activity, showed typically seasonal variations. Validity was established by correlating physical activity data with maximum oxygen uptake. Maximum oxygen uptake correlated with sport activities (partial correlation coefficient: r = 0.422, p < 0.01). Evaluated data were consistent. People rating themselves as "more active than their coevals" were indeed more active in sport (r = 0.334, p < 0.01) and total activity (r = 0.282, p < 0.05). Studying activity patterns of a population sample of adult residents of Freiburg (systematic random sampling, n = 612, 20-98 years) we found total physical activity of 9.2 hours per week (median), with activities of low to moderate intensities dominating. Age and gender are important determinants of the activity patterns. According to the recommendation of Paffenbarger (2000 kcal/week total physical activity) 40% of the residents of Freiburg did not reach the recommended energy expenditure. Compared to the recommendation of the American College of Sports Medicine (1000 kcal/week by training) 63% of the population sample were not active enough.