The pressure-flow relationship has been studied in a peripheral portion of the lung vasculature in anesthetized dogs with use of a double-lumen catheter wedged in a distal pulmonary artery. One lumen was used to infuse mixed venous blood in the wedged area and the other to measure the corresponding perfusion pressure. Flow ranged from 0 to 9.2 ml/min, and the mean volume of the wedged area (n = 59) was 0.75 +/- 0.05 (SE) ml. In the areas where the distal pulmonary artery was in the same direction as the catheter ("coaxial"), the mean pressure-flow curve showed a negligible gamma-intercept and no significant difference between ascending and descending flow. The slope of the initial part of the ascending limb (peripheral pulmonary vascular resistance) varied from site to site and did not show a significant correlation with the overall pulmonary vascular resistance; it was inversely correlated with the volume of the wedged area (r = -0.35, P < 0.05) and directly, as expected, correlated with the y-intercept (r = 0.78, P < 0.001) and hysteresis (r = 0.48, P < 0.001). The results of two consecutive pressure-flow runs in the same site showed similar results, with no difference exceeding the error of measurement. In contrast, the slope increased by 71% during hypoxia (fraction of inspired O2 was 0.10, n = 5). This procedure seems suitable to determine the effects of physiological or pharmacological interventions on the pulmonary vessels, without interference of the systemic circulation.