The period prevalence of simple chronic bronchitis (SCB) (mucus hypersecretion), defined as chronic cough and sputum production by the MRC respiratory symptom questionnaire administered by occupational physicians and of obstructive chronic bronchitis (OCB) (airflow obstruction) (defined as SCB plus FEV1 less than 80% predicted) have been measured over the period 30 June 1977-30 June 1980 in the entire work force aged between 21 and 60 of the coal industry of New South Wales, Australia (12 357 men). Four dimensional contingency table analysis by a logistic transform method showed highly significant (p less than 0.001) additive affects of age (exposure duration), site of work, smoking, and alcohol consumption on development of overall chronic bronchitis (SCB + OCB). Odds ratios were face work:surface work = 1.78:1, smoker:non-smoker = 4.23:1, alcohol greater than 300 g/wk:alcohol less than 300 g/wk = 2.13:1. There was no evidence for synergistic effects of these factors on the development of mucus hypersecretion. When OCB was analysed separately, the effect of site of work, although in the same direction, was not statistically significant and this was assumed to be due to a "healthy worker" effect or a "swamping" effect of smoking. Age, smoking, and alcohol effects were highly significant (p less than 0.0001) and there was a sharp increase in prevalence of OCB in the age groups 41-50 and 51-60. Odds ratios were face work:surface work = 1.11:1, smoker:non-smoker = 2.66:1, alcohol greater than 300 g/wk:alcohol less than 300 g/wk = 2.91:1. There was no evidence of synergistic effects. These results are consistent with a hypothesis of additive effects of smoking, alcohol, and coal mine dust and fumes on the development of chronic mucus hypersecretion leading to airflow obstruction or a hypothesis of similar additive effects on the development of two separate conditions--mucus hypersecretion with airflow obstruction and mucus hypersecretion without airflow obstruction.