There are two fundamentally different methods for selecting appropriate general circulation and regional climate models (GCMs/RCMs). The first approach aims to cover the full envelope of possible futures ranging from dry and cold projections to wet and warm projections, while the second approach selects GCMs/RCMs on the basis of indicators of past performance. In the case of the Hindu Kush Himalayas (HKH), the first approach may be preferable as climate models have considerable difficulty in simulating past climate. In this study, we develop a new method that combines the two existing methods, so as to have a set of climate models that not only cover a wide range of possible futures but also reproduce the most important processes in the region.