We studied the effect of intravenous dextran 70 infusion on lung liquid and protein exchange to determine whether its effects were due to altered hemodynamics or to altered microvascular permeability. In each of six instrumented awake sheep with chronic lung lymph fistulas, we performed three experiments: 1) control, 2) a 30-minute infusion of 1 l of 6% dextran 70, and 3) an infusion of 1 l of 0.9% NaCl. In addition to pulmonary hemodynamics and lymph dynamics, we measured the plasma-to-lung lymph equilibration rate of [125I]albumin. We followed all the sheep for 10 hours, including a 2-hour baseline period. Dextran was more effective in expanding plasma volume (63 ± 15% [mean ± SD]) than saline (11 ± 6%) at the end of the 30-minute infusion. Pulmonary vascular pressures increased after dextran and remained elevated for 8 hours, whereas after saline the pressures returned to baseline within 1 hour. After dextran, lung lymph flow increased and remained elevated. It was only transiently increased after saline. We confirmed that dextran equilibrated rapidly with lung lymph (half-time, <0.6 hour), even though it maintained plasma volume expansion for the whole body (half-time, 11.1 ± 2.7 hours). The dextran increased both plasma and lymph total macromolecular osmotic pressure but did not increase the plasma-interstitial (lymph) osmotic pressure difference in the lung, except transiently during the infusion. The lymph/plasma protein concentration ratio increased after dextran due mainly to plasma protein dilution. There were no differences in the half-time of tracer albumin equilibration between plasma and lung lymph (control, 2.2 ± 0.6 hours; saline, 2.0 ± 0.6 hours; dextran, 2.3 ± 0.6 hours). Dextran 70 increased liquid filtration mainly by increasing microvascular pressure and possibly filtration surface area. There was no evidence for a change in the leakiness of the lung microvascular barrier to albumin.