Climate projections for the next 20-50 years forecast higher temperatures and variable precipitation for many landscapes in the western United States. Climate changes may cause or contribute to threshold shifts, or tipping points, where relatively small shifts in climate result in large, abrupt, and persistent changes in landscape patterns and fire regimes. Rather than simulate potential climate - fire interactions using future climate data derived from Global Climate Models (GCMs), we developed sets of progressively warmer and drier or wetter climate scenarios that span and exceed the range of GCM outputs for the western US, including temperature and precipitation combinations that may not be present in GCM projections but may occur at finer (regional or local) scales. These climate scenarios were used to simulate potential future fire and vegetation dynamics in three study areas in the western United States - McDonald watershed, Glacier National Park (MT), the central plateau of Yellowstone National Park (WY), and the East Fork Bitterroot River basin (MT). These landscapes encompass a diverse range of biophysical settings, vegetation species, forest structure, and fire regime, and thus were expected to differ in their sensitivity to climate changes and exhibit unique threshold behavior following climatic and wildfire perturbations. Each of the study areas proved sensitive to simulated changes in temperature and precipitation, as reflected in shifts in mean annual burned area, crown fire area, and fire - caused tree mortality. Sensitivity to climate changes differed across Landscapes - for example, a significant decline in basal area occurred at temperature shifts of 3 °C and above for the Yellowstone National Park study area, 4 °C and above for the Glacier National Park study area, and above 5 °C for the East Fork Bitterroot River basin. Moreover, shifts in basal area were strongly related to changes in area burned and fire regime characteristics, suggesting that synergistic interactions of climate and fire will be important in determining future landscape patterns.