In the summer of 1996 the Department of Energy G-1 aircraft was deployed in the New York City metropolitan area as part of the North American Research Strategy for Tropospheric Ozone-Northeast effort to determine the causes of elevated O3 levels in the northeastern United States. Measurements of O3, O3 precursors, and other photochemically active trace gases were made upwind and downwind of New York City with the objective of characterizing the O3 formation process and its dependence on ambient levels of NOx and volatile organic compounds (VOCs). Four flights are discussed in detail. On two of these flights, winds were from the W-SW, which is the typical direction for an O3 episode. On the other two flights, winds were from the NW, which puts a cleaner area upwind of the city. The data presented include plume and background values of O3, CO, NOx, and NOy concentration and VOC reactivity. On the W-SW flow days O3 reached 110 ppb. According to surface observations the G-1 intercepted the plume close to the region where maximum O3 occurred. At this point the ratio NOx/NOy was 20--30%, indicating an aged plume. Plume values of CO/NOy agree to within 20% with emission estimates from the core of the New York City metropolitan area. Steady state photochemical calculations were performed using observed or estimated trace gas concentrations as constraints. According to these calculations the local rate of O3 production P(O3) in all four plumes is VOC sensitive, sometimes strongly so. The local sensitivity calculations show that a specified fractional decrease in VOC concentration yields a similar magnitude fractional decrease in P(O3). Imposing a decrease in NOx, however, causes P(O3) to increase. The question of primary interest from a regulatory point of view is the sensitivity of O3 concentration to changes in emissions of NOx and VOCs. A qualitative argument is given that suggests that the total O3 formed in the plume, which depends on the entire time evolution of the plume, is also VOC sensitive. Indicator ratios O3/NOz and H2O2/NOz mainly support the conclusion that plume O3 is VOC sensitive. ¿ 2000 American Geophysical Union