OBJECTIVE To study the role of dust exposure on incidence of respiratory symptoms and decline of lung function in young coal miners. METHODS The loss of lung function (forced vital capacity (FVC), forced expiratory volume in one second (FEV1), forced expiratory flow (MEF), carbon monoxide transfer factor (TLCO)) with time and the incidence of respiratory symptoms in 909 Sardinian coal miners (followed up between 1983 and 1993 with seven separate surveys) has been compared with the past and current individual exposures to respirable mixed coal dust. Multiple linear and logistic regression models were used simultaneously controlling for age, smoking, past occupational exposures, and other relevant covariates. RESULTS According to the relatively low dust exposures experienced during the follow up few abnormal chest x ray films were detected. In the cross sectional analysis of initial data, significant associations between individual cumulative exposure to dust, decrements in FEV1 and MEFs, and increasing prevalence of respiratory symptoms were detected after allowing for the covariates included in the model. The yearly decline of FVC, FEV1, and single breath carbon monoxide transfer factor (TLCO/VA) was still significantly related to the individual exposure to dust experienced during the follow up, even after allowing for age, smoking, initial cumulative exposure to dust, and initial level of each functional variable. In logistic models, dust exposure was a significant predictor of the onset of respiratory symptoms besides age and smoking. CONCLUSIONS The results show that even moderate exposures to mixed coal dust, as in our study, significantly affect lung function and incidence of symptoms of underground miners. Although the frequency of chest x ray examination might be fixed at every three or four years, yearly measurements of lung function (spirometry, MEFs, and TLCO) are recommended for evaluation of the respiratory risk from the coal mine environment to assess the need for further preventive interventions.