Apparently conflicting differences between the regional chest wall motion and gas transport have been observed during high-frequency ventilation (HFV). To elucidate the mechanism responsible for such differences, a positron imaging technique capable of assessing dynamic chest wall volumetric expansion, regional lung volume, and regional gas transport was developed. Anesthetized supine dogs were studied at ventilatory frequencies (f) ranging from 1 to 15 Hz and eucapnic tidal volumes. The regional distribution of mean lung volume was found to be independent of f, but the apex-to-base ratio of regional chest wall expansion favored the lung bases at low f and became more homogeneous at higher f. Regional gas transport per unit of lung volume, assessed from washout maneuvers, was homogeneous at 1 Hz, favored the bases progressively as f increased to 9 Hz, and returned to homogeneity at 15 Hz. Interregional asynchrony (pendelluft) and right-to-left differences were small at this large regional scale. Analysis of the data at a higher spatial resolution showed that the motion of the diaphragm relative to the excursions of the rib cage decreased as f increased. These differences from apex to base in regional chest wall expansion and gas transport were consistent with a simple model including lung, rib cage, and diaphragm regional impedances and a viscous coupling between lungs and chest wall caused by the relative sliding between pleural surfaces. To further test this model, we studied five additional animals under open chest conditions. These studies resulted in a homogeneous and f-independent regional gas transport. We conclude that the apex-to-base distribution of gas transport observed during HFV is not caused by intrinsic lung heterogeneity but rather is a result of chest wall expansion dynamics and its coupling to the lung.