In this article, a risk based approach to design for bushfire protection in view of adaptation to global warming is discussed. The concept of design bushfire is explained in an analogy to design flood or design earthquake in terms of event of prescribed return period. In lieu of using the Global Climate Model, the current study is based on the analysis of historical fire weather data from multiple locations in a state wide region. The generalized extreme value (GEV) analysis method is employed to establish the recurrence models for predicting the fire weather index of given return period and the associated fire intensity. To examine the impacts of the climate change, a moving GEV method is utilized to the weather data records over the period of 44 years. The result demonstrated a heterogeneity in the impact of climate change in terms of a given recurrence fire danger index and the potential bushfire severity over the region studied. The implication of this outcome is that the traditional prescriptive approach to design for bushfire protection may not be suited for adaptation to climate change.