Objectives: To evaluate and compare physical activity patterns of urban and rural dwellers in Cameroon, and study their relationship with obesity, diabetes and hypertension., Methods: We studied 2465 subjects aged >or=15 y, recruited on the basis of a random sampling of households, of whom 1183 were urban dwellers from Yaoundé, the capital city of Cameroon and 1282 rural subjects from Bafut, a village of western Cameroon. They all had an interviewer-administered questionnaire for the assessment of their physical activity and anthropometric measurements, blood pressure and fasting blood glucose determination. The procedure was satisfactorily completed in 2325 (94.3%) subjects. Prevalences were age-adjusted and subjects compared according to their region, sex and age group., Results: Obesity was diagnosed in 17.1 and 3.0% urban and rural women, respectively (P<0.001), and in 5.4 vs 1.2% urban and rural men, respectively (P<0.001). The prevalence of hypertension was significantly higher in urban vs rural dwellers (11.4 vs 6.6% and 17.6 vs 9.1% in women and men, respectively; P<0.001). Diabetes was more prevalent in urban compared to rural women (P<0.05), but not men. Urban subjects were characterized by lower physical activity (P<0.001), light occupation, high prevalence of multiple occupations, and reduced walking and cycling time compared to rural subjects. Univariate analysis showed significant associations between both physical inactivity and obesity and high blood pressure. The relationship of physical inactivity with hypertension and obesity were independent in both urban and rural men, but not in women. Body mass index, blood pressure and glycaemia were higher in the first compared with the fourth quartiles of energy expenditure., Conclusion: Obesity, diabetes and hypertension prevalence is higher in urban compared to rural dwellers in the populations studied. Physical activity is significantly lower and differs in pattern in urban subjects compared to rural. Physical inactivity is associated with these diseases, although not always significant in women.