This study proposes a probabilistic drought forecasting model to forecast meteorological drought in Southern Taiwan using the El Niño-Southern Oscillation (ENSO) index. Meteorological drought is defined by the standardized precipitation index (SPI), and the ENSO index is El Niño sea surface temperature (SST). Two probabilistic forecasting model architectures were constructed based on the transition probabilities from El Niño SSTs to SPIs. Both model architectures forecast a one-month-ahead probability distribution for meteorological drought using different combinations of El Niño SST variables. Forecasting results showed the robustness of the probabilistic drought forecasting models. In addition, this study discussed the selection of El Niño SST variables used in the probabilistic drought forecasting model, and found that models with a single SST input outperformed those with multiple SST inputs.