We examined the effects from subjects, technicians and spirometers on within-session variability in successful recordings of forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC) in 4989 asymptomatic never-smoking men. All eligible men aged 30-46 years living in western Norway (n = 45,380) were invited to a cross-sectional community survey. Information on respiratory symptoms, smoking habits and occupational exposures was obtained from a self-administered questionnaire. Three successful FEV1 and FVC recordings were obtained in 26,368 attendants using three dry-wedge bellow spirometers operated by 10 different technicians. Within-subject standard deviation (SD) from three recordings of FEV1 and FVC was on average 102 and 106 ml, respectively, and increased with height (14 and 17 ml, respectively, per 10 cm) and body mass index (BMI) (11 and 14 ml, respectively, per 5 kg m-2). Between-subject SD of the mean of three FEV1 and FVC recordings was 591 and 754 ml, respectively, and increased in groups of increasing height (43 and 40 ml, respectively, per 10 cm). Small, but significant, differences were observed between technicians in within-subject SD and in levels of FEV1 and FVC. Homogeneity of between-subject variability, necessary for linear regression analysis, was obtained using FEV1 and FVC divided by height squared. In conclusion, within-subject variability in three successful spirometric recordings was small, but dependent on height and BMI of the subjects as well as technician performance. The observed heterogeneity in between-subject variation in FEV1 and FVC levels disappeared when each variable was divided by height squared. Novel multiple linear regression equations for FEV1/height2 and FVC/height2 were developed to be used in evaluating the effects from occupational airborne exposures in Nordic men aged 30-46 years.