The variability (coefficient of variation of five consecutive measurements), reproducibility (difference of results at 1 h and 24 h), and interobserver difference (independent reading of the tracings by two observers) of airways resistance (Raw) and static lung volumes (residual volume, functional residual capacity, total lung capacity) using a body plethysmograph were assessed in 14 healthy subjects and in 25 patients with various respiratory disorders. The variability was low for TLC (4-5%), moderate for FRC (7-8%) and high for Raw (28%). No significant changes of Raw or lung volumes were found for the groups at 1 h and 24 h. Between observers, a slight difference existed for FRC and Raw in normal subjects; the difference was higher (4.5% for FRC and 11% for Raw) and became significant in patients. The overestimation of Raw by observer 2 as compared to observer 1 was more important at larger values. The present findings call for caution when pooling results obtained by several observers in large-scale studies, or when comparing figures obtained by different technicians in the pulmonary function laboratory.