To assess the actual sensitivity of forced expiration indices to changes in the resistance of peripheral airways, maximal expiratory flow-volume curves were obtained in 30 normal subjects breathing air and breathing an O2-Ne-SF6 gas mixture having the same density as air but a 45% larger viscosity. The measurements were made using a bag-in-box system to circumvent calibration problems, and the data were processed digitally. Besides the usual forced expiration indices, slope ratios, as described by Mead (J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 44: 156-165, 1978), and transit time indices were also computed. Breathing the viscous mixture, the largest changes were seen with the slope ratio measured at 60% forced vital capacity, followed by other slope ratios, the standard deviation of transit times, and maximal flows at low lung volumes. However, when the data were compared using the t test for paired measurements, the most significant changes were found with the forced expiratory volumes at 1 and 2 s, due to their low within-subject variability. These indices may therefore by considered as the most suitable for detecting changes in individuals.