Only limited data are available about the incidence of asthma based on longitudinal prospective studies. Further, the results from different studies on incidence vary considerably depending on the age composition of the cohorts under study, the used methods and the criteria for disease. Also among adults high incidence rates have been reported during recent years. The aim of this study was to examine to what extent the incidence of physician-diagnosed asthma could be explained by a real incidence of the disease, and to what extend by an increased diagnostic activity or altered diagnostic praxis. Another aim was to study risk factors for asthma based on incident cases. Three cross-sectional surveys have been performed in the same population sample living in the northern-most province of Sweden, Norrbotten. The first survey was performed in 1986, and 5698 subjects, 86% of those invited, responded to a postal questionnaire. Of these, 4754 subjects (83%) participated at the third survey in 1996. After exclusion of all subjects who had reported that they had asthma in 1986, or had been classified as having asthma in 1986, 68 men and 98 women (P=0.02) reported in 1996 that they had been diagnosed as having asthma by a physician. Thus, the cumulative incidence for the 10-year period was 3.2% among men and 4.5% among women. After correction for subjects who already in 1986 had reported symptoms common in asthma, or had been classified as having chronic bronchitis, 97 subjects with incident asthma remained, which corresponded to an annual incidence rate among men of 1.7 and among women of 2.9/1000 persons year(-1) (P=0.1). Clinical examinations confirmed asthma in a large majority of these 97 subjects. Significant risk factors were family history of asthma, both ex- and current smoking, and female sex. The socio-economic groups manual workers and assistant non-manual employees were associated with incident asthma, although not significantly. The increasing prevalence of asthma among adults during recent 10-20 years may to a considerable extent be explained by an increased diagnostic activity or altered diagnostic praxis. Use of different methods when measuring incidence may in part explain the extremely diverging incidence rates of asthma found in different studies.