A cross sectional study of risk factors in respiratory diseases was carried out in August 1986, in Beijing, China. Inhabitants greater than or equal to 40 years old were selected at random from a rural area, a residential area and an industrial area, using a two stage sampling method. The analysis presented here is based on the sample population of adults who (1) were prepared to be interviewed, using the British Medical Research Council's questionnaire translated into Chinese (n = 3423) and (2) had lung function measurements at the same time (n = 3373). Obstructive lung disease was defined as forced expiratory volume in 1s (FEV1) less than 68% of forced vital capacity (FVC). Seven variables were considered as potential risk factors or confounding factors: area of residence, sex, age, cigarette smoking, history of respiratory disease, socio-economic status and familial component. A modified binary variable regression method developed by Feldstein was used for the adjustment of rate ratios. The adjusted prevalence of obstructive lung disease was highest in the rural area and lowest in the residential area(s). An increase in age, cigarette smoking, low socio-economic status and positive history of respiratory diseases were associated with significantly higher rates of impaired pulmonary function. The other measured factors did not appear to be related to impaired pulmonary function.