Knowledge of the factors determining the performance of bronchodilator reversibility testing in a general population is lacking. Random samples of all adults aged 47-48 and 71-73 years living in Bergen, Norway, were invited to a cross-sectional study. Altogether 3506 subjects attended (69%). Test quality predictors were examined through multiple logistic regression analyses including gender, age, body mass, educational level, smoking history, respiratory symptoms, and in the elderly cohort cognitive level. Among the participants, 1.6% refused to inhale salbutamol, 2.5% failed the initial spirometry according to the ATS guidelines, and 1.3% failed the post-bronchodilator spirometry. Old age and body mass index > 30 kg/m2 were independent risk factors for an unsuccessful initial spirometry, and never smoking and respiratory symptoms were risk factors for failing the post-bronchodilator spirometry. Cognitive impairment in the elderly was a risk factor for failing both the initial- and post-bronchodilator spirometry. The median number of forced expirations was 7 in subjects obtaining an acceptable reversibility test. One third of these participants needed > or = 8 attempts, with independent predicting factors being old age, Little formal education and never smoking. Although reversibility testing becomes increasingly difficult with age, reliable data are obtained in a vast majority (94%) of subjects in our community study.