Inhaled nitric oxide (NO) has been reported to improve oxygenation in patients with COPD if administered in combination with oxygen (O2). Little, however, is known about the variability of these effects and the potential influence of body position. Twenty-six spontaneously breathing patients with moderate to severe COPD inhaled clean air, O2(FiO2, 0.29), 5 ppm NO, 5 ppm NO+O2, 10 ppm NO+O2, 10 ppm NO, and again clean air in an upright position. Blood gas analysis from arterialized capillary blood was performed after each inhalation. Tests were repeated on different days to assess the variability of the response. Furthermore, eight patients were studied in both upright and supine position while inhaling 5 ppm NO in the presence or absence supplemental O2. As compared to clean air, NO led to a mean decrease in PaO2 of -0.9 mmHg at 5 ppm and of -2.8 mmHg at 10 ppm NO. Similarly, NO+O2 led to a dose-dependent fall in PaO2 of -1.8 and -3.6 mmHg, respectively, as compared to O2. Average within-subject variation (SD) of the effects elicited by 5 and 10 ppm NO was 2.4 and 2.3 mmHg without additional O2, and 4.7and 5.3 mmHg with O2. The effects of 5 ppm NO+O2 differed significantly between upright and supine position; as compared to O2 alone, mean (SD) changes were -3.7 +/- 5-8 vs. +1.1 +/- 4.9 mmHg, respectively. Our findings suggest thatthe addition of NO to inhaled oxygen, when given in an upright position, does not lead to an improvement of PaO2 in patients with moderate to severe COPD. Furthermore, it turned out that it was not possible to define responders and non-responders to inhaled NO on an individual basis, since the variability ofthe responses was similar to the mean