The launching of cheap, pocket-sized spirometers, with data storage capability, has made patient-administered sequential spirometry (PASS) an attractive method of monitoring ventilatory capacity. At present, little information is available on the quality of PASS, compared to laboratory spirometry. The aim of this study was to investigate whether patients could perform PASS without loss of reliability and reproducibility as compared with traditional laboratory spirometry. Ten healthy volunteers performed spirometry for 1 month and 10 emphysematous patients with alpha 1-antitrypsin deficiency (type PiZ) performed spirometry twice daily for up to 2 yr. To fulfil Good Clinical Practice criteria on full data documentation, a traditional direct recording spirometer, the Vitalograph R-model, was used. A decompression device was used for calibration and a 3.8% annual drift in volume registration was noted. This drift was largest for the first year. After training, all patients were able to perform unsupervised spirometry, producing technically correct forced expiratory curves. Reproducibility of FEV1 and FVC obtained by PASS was found to be as good as for laboratory spirometry. After adjustment for the diurnal variation, the residual variation of FEV1 was 2.5% (range 1.6-4.2%) for healthy volunteers and 5.6% (range 4.2-7.7%) for emphysematous patients. Forced vital capacity showed the same pattern. In conclusion, PASS is possible in highly motivated individuals without loss of reliability and reproducibility when compared to laboratory spirometry.