We measured thoracic gas volume (Vtg) by the body plethysmographic technique in 4 normal subjects and in 9 asthmatics during remission and after bronchoconstriction induced with inhaled aerosolized histamine. Changes in both mouth (Pm) and esophageal pressure (Pes) were used to approximate changes in alveolar pressure (Palv), and both Vtg(m) and Vtg)(es) were computed. Because small changes in lung recoil occur during panting, delta Pes exceeds delta Pm, leading to a value of Vtg(m) that is slightly greater than Vtg(es). Indeed, in the normal subjects, Vtg(m) was significantly greater than Vtg(es), and the observed percentage error (E) between them (2.8%) corresponded with the expected value. In the asthmatics, during marked bronchoconstriction, there was an increased tendency for Vtg(m) to exceed Vtg(es), E averaging 12.2% and being greater than the predicted error. In some asthmatics, changes in volume led changes in Pm; the phase angle between volume and Pm averaged 6.3 degrees and was significantly greater than that between volume and Pes (1.7 degrees). In 3 asthmatics, increases in TLC(m) were not accompanied by similar changes in TLC(es). Our data are consistent with the hypothesis that when a compliant upper airway is coupled with a high airway resistance during panting against an occluded airway, transmission of changes in alveolar pressure to the mouth is incomplete. The resulting underestimation of delta Palv by delta Pm leads to artifactual increases in TLC.