To assess the effect of pregnancy, maternal position, and cardiac output on intrapulmonary shunting (Qs/Qt) in normotensive nulliparous women near term. Ten normotensive nulliparas between 36 and 38 weeks' gestation underwent pulmonary artery catheterization (via the subclavian route) and radial artery canalization. Baseline assessments were made with subjects in the left lateral recumbent position after a 30-minute stabilization period. Measurements were obtained sequentially in the left lateral, right lateral, supine, knee-chest, sitting, and standing positions. Each position change was followed by a 10-minute pre-measurement stabilization period. Cardiac output was measured via the thermodilution technique. Blood samples were obtained simultaneously from the pulmonary and radial arteries and analyzed in duplicate for oxygen content with a blood gas analyzer. Qs/Qt was calculated using the classic shunt equation. Statistical analysis was performed by analysis of variance of repeated measures of Qs/Qt and maternal position. The relationship of Qs/Qt to maternal cardiac output was evaluated by the correlation coefficient. Significance was defined as P < .05. Directly measured Qs/Qt averaged 15.3% in left lateral, 15.2% in right lateral, 13.9% in supine, 12.8% in knee-chest, 13.8% in sitting, and 13.0% in standing positions. There was no statistically significant correlation between Qs/Qt and cardiac output (R2 = 0.11, not significant). This is the first report of directly measured Qs/Qt in normal pregnant women in the third trimester. Qs/Qt values reported in pregnancy are higher than those reported in nonpregnant individuals.