AbstractThe influence of Madden‐Julian Oscillation (MJO) on the rainfall distribution of Southeast Asia is studied using TRMM satellite‐derived rainfall and rain gauge data. It is shown that convectively active (suppressed) phases of MJO can increase (decrease) the probability of extreme rain events over the land regions by about 30–50% (20–25%) during November–March season. The influence of MJO on localized rainfall extremes are also observed both in rainfall intensity and duration. The Met Office Global Seasonal forecasting system seasonal forecasting system is shown to reproduce the MJO influence on rainfall distribution well despite the model biases over land. Skills scores for forecasting 90th percentile extreme rainfall shows significant skills for convective phases. This study demonstrates the feasibility of deriving probabilistic forecasts of extreme rainfall at medium range.