Because forest fires emit substantial NOx and hydrocarbons--known contributors to O3 production--we hypothesize that interannual variation in western U.S. O3 is related to the burned area. To evaluate this hypothesis we used a gridded database of western U.S. summer burned area (BA) and biomass consumed (BC) by fires between 101-125 degrees W. The fire data were compared with daytime summer O3 mixing ratios from nine rural Clean Air Status and Trends Network (CASTNET) and National Park Service (NPS) sites. Large fire years exhibited widespread enhanced O3. The summer BA was significantly correlated with O3 at all sites. For each 1 million acres burned in the western U.S. during summer, we estimate that the daytime mean O3 was enhanced across the region by 2.0 ppbv. For mean and maximum fire years, O3 was enhanced by an average of 3.5 and 8.8 ppbv, respectively. At most sites O3 was significantly correlated with fires in the surrounding 5 x 5 degrees and 10 x 10 degrees regions, but not with fires in the nearest 1 x 1 degree region, reflecting the balance between O3 production and destruction in a high NOx environment. BC was a slightly better predictor of O3, compared with BA. The relationship between O3 and temperature was examined at two sites (Yellowstone and Rocky Mountain National Parks). At these two sites, high fire years were significantly warmer than lowfire years; however, daytime seasonal meantemperature and O3 were not significantly correlated. This indicates that the presence of fire is a more important predictor for O3 than is temperature.