To examine how molecular charge affects the transfer of molecules across the alveolar-capillary barrier, we prepared the following dextrans of equivalent molecular size (mol wt 10,000) but varying molecular charge: neutral dextran, cationic DEAE dextran, and anionic dextran sulfate. These were labeled with 99mTc. The lungs of three groups of anesthetized rabbits were insufflated with dextran aerosols, with six rabbits receiving each type, and the half-time pulmonary clearance (t1/2) was measured. Control t1/2's (95% confidence limits) were 95 (74-120), 227 (192-268), and 291 (246-345) min for neutral, cationic, and anionic dextrans, respectively. One week later, when the same animals were restudied 4 h after 3 micrograms/kg iv endotoxin, t1/2's were 102 (75-139), 167 (149-187), and 126 (102-154) min, respectively. After 30 min during this repeat study, animals were ventilated with 20 breaths of cigarette smoke, which acutely increased the clearance rate to 34 (26-46), 25 (20-31), and 13 (7-24) min, respectively. Mean carboxyhemoglobin levels were not significantly different in the three groups: 13.6, 12.7, and 11.1%, respectively. These results demonstrated that neutral dextrans showed the same clearance rate before and after endotoxin, whereas the charged dextrans had a significantly faster clearance after endotoxin. After smoke exposure the anionic dextran left the lung more rapidly than the neutral dextran. Thus molecular charge affects solute transfer across the alveolar-capillary barrier in both normal and injured lungs, and an effect of endotoxin on the lung can be detected with charged dextrans but not with neutral dextran.