Seasonal mean rainfall projections for Hawai'i are given based on statistical downscaling of the latest Coupled Model Intercomparison Project phase 5 (CMIP5) global model results for two future representative concentration pathways (RCP4.5 and RCP8.5). The spatial information content of our statistical downscaling method is improved over previous efforts through the inclusion of spatially extensive, high-quality monthly rainfall data set and the use of improved large-scale climate predictor information. Predictor variables include moisture transport in the middle atmosphere (700 hPa), vertical temperature gradients, and geopotential height fields of the 1000 and 500 hPa layers. The results allow for the first time to derive a spatially interpolated map with future rainfall change estimates for the main Hawaiian Islands. The statistical downscaling was applied to project wet (November-April) and dry (May-October) season rainfall anomalies for the middle and late 21st century. Overall, the statistical downscaling gives more reliable results for the wet season than for the dry season. The wet-season results indicate a pronounced dipole structure between windward facing mountain slopes and the leeward side of most of the islands. The climatically wet regions on the windward slopes of the mountain regions are expected to become wetter or remain stable in their seasonal precipitation amounts. On the climatically dry leeward sides of Kaua'i, O'ahu, Maui, and Hawai'i Island, future precipitation exhibits the strongest drying trends. The projected future rainfall anomaly pattern is associated with a circulation anomaly that resembles a shift in the position or strength of the subtropical high and the average location of extratropical troughs. These new results suggest that a negative trend dominates the area-averaged changes in the statistical downscaling over the Hawaiian Islands. However, the islands are expected to experience a greater contrast between the wet and dry regions in the future.