In this study, an ensemble system is developed using the WRF model to produce probabilistic precipitation forecasts over Iran. The ensemble system consists of WRF model simulations with eight different physical configurations. Initial and boundary conditions for WRF are provided from the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) forecasts with a horizontal resolution of 0.5. The Bayesian model averaging (BMA) method is used to calibrate the probabilistic forecasts of rainfall in the fall and winter of 2016-2015. The 24-, 48- and 72-hour raw ensemble outputs were calibrated using the BMA technique for 90 days as the test period with 87 training days of forecasts. The calibrated probabilistic forecasts are assessed using reliability diagram, ROC diagram, Brier score and RPS. In addition, the forecast economic value has been investigated. The results show that the application of BMA has improved the reliability of the raw ensemble, such that the reliability and ROC diagrams have been improved significantly. For 24-hour forecasts, the Brier score is reduced by 24, 30, 32, 36, 39 and 65 percent at the thresholds of 0.1, 2.5, 5, 10, 15 and 25 mm, respectively. Similar results were achieved in 48- and 72-hour forecasts. The RPS score for the 24-, 48- and 72-hour forecasts is reduced by 45, 40 and 38 percent, respectively.