Measurements of NO, NOy, O3, and CO were made during NASA's Global Tropospheric Experiment/Pacific Exploratory Mission-West B (GTE/PEM-West B) carried out over the western Pacific in February and March 1994. NOx was calculated from NO using a photostationary state model ((NOx)mc). Correlations between these species are presented, and some insights into the sources of NOx and NOy are described. The boundaries between the lower, middle, and upper troposphere have been defined at potential temperatures of 311 K and 328 K, which correspond to the geometric altitudes of about 5 and 9 km at 30 ¿N. Enhancements in the mixing ratios of NOy and CO were observed in the lower and middle troposphere. A positive correlation was found between these two species suggesting that the high NOy values were due to anthropogenic emissions over the continental surface. On the other hand, O3 increased little with increase in CO. As a result, NOy/O3 ratios were higher in air more influenced by pollution. NOy values in 55 and 28% of the air masses sampled in the lower and middle troposphere, respectively, were higher than the clean free tropospheric NOy-O3 range when O3 values simultaneously observed were used. High (NOx)mc/NOy ratios between 0.15 and 0.3 were found in the boundary layer with relatively low mixing ratios of CO and NOy during the three flights. These air masses were transported from a higher altitude (~5 km) and a higher latitude (~50 ¿N) within a few days. The peroxyacetyl nitrate (PAN)/NOy ratios were generally high (~0.4) in these air masses, and the thermal decomposition of PAN was a probable source of NOx. In the middle troposphere the (NOx)mc mixing ratio did not generally increase with NOy or CO, suggesting that the transport of air masses affected by anthropogenic emissions did not increase the NOx level significantly. In the upper troposphere, very minor effects from the continental surface sources were seen in the CO mixing ratio. By contrast, NOy values in 33% of the air masses were higher than those expected when stratospheric air intrusion is assumed to be a single source of NOy based on NOy-O3 correlation analyses. This result suggests significant free tropospheric NOy sources, namely exhaust from the aircraft and NO production by lightning activity. In fact, spikes in the (NOx)mc mixing ratios were observed near the aircraft corridor south of Tokyo at an altitude of 10 km. These two free tropospheric NOx sources were considered to be important in determining the levels of the upper tropospheric NOx and NOy during PEM-West B. ¿ 1997 American Geophysical Union