Smoke haze from forest fires is among Southeast Asia’s most serious environmental problems and there is a clear need for a fire and haze early warning system (EWS) for the region. APCC has been collecting monthly dynamic prediction data produced by 16 institutions and has been producing 6-month lead Multi-Model Ensemble (MME) climate forecasts every month. In this study, we developed 4 different statistical downscaling methods and assessed the forecast skill of each method over fire-prone regions in Southeast Asia. We developed a EWS prototype in which 3-month precipitation (August to October) is predicted during April to July and the forecasted precipitation amount is then translated into four fire danger ratings based on the relationship between precipitation amount and CO2 emission. A needs assessment for early warning Information was conducted through the field survey with resource managers at three provinces in Indonesia. A two day workshop was held at the Malaysian Meteorological Department (MMD) with financial and logistical support from MMD for the improvement of the EWS. Finally, the forest fire early warning information on Southeast Asia created using the EWS will be provided though the hosting server in APCC.