This paper focuses upon flow-resistive pulmonary work during upright immersion, and during changes in the air delivery pressure. Nine male non-smokers (aged 26.2 +/- 3.5 years), with normal lung function history, performed spontaneous respiration while seated in air (control) and during total immersion. During the immersed state subjects were supplied with air at four hydrostatic pressures: mouth pressure (PM; simulating a mouth-held demand regulator), lung centroid pressure (PLC; + 1.33 kPa relative to the sternal notch), and 0.98 kPa (10 cmH2O) above and below the lung centroid pressure. Inspiratory, expiratory and total flow-resistive pulmonary work were computed from the integration of transpulmonary pressure (difference between oesophageal and mouth pressure) with respect to lung volume change. When breathing air delivered at mouth pressure, immersion significantly elevated all total flow-resistive pulmonary work components (P less than 0.05). Each increment in breathing pressure resulted in a progressive reduction in expiratory and total flow-resistive pulmonary work, so that when air was provided at lung centroid pressure and lung centroid pressure +0.98 kPa both components were similar to control values (P greater than 0.05). Inspiratory was always less than expiratory pulmonary work. During immersion inspiratory pulmonary work was significantly reduced when air supply pressure was increased above mouth pressure (P less than 0.05). Subsequent pressure increments failed to produce further changes in inspiratory pulmonary work. The difference in response between the inspiratory and expiratory components of total flow resistive pulmonary work was attributed primarily to the volume-dependence of the expiratory component.