For a large variety of applications, it is crucial to be able to predict the electric range of a vehicle in a robust and reliable way. Current applications show a lack of accuracy and robustness. The designed algorithm is capable of accurately predicting the range in a robust way, incorporating the effects of drive style, type of route, type of vehicle and some weather effects. The algorithm is model based and uses a power consumption model, battery model and drive cycle model to determine the remaining e-range of an (H)EV. Some model parameters, which are crucial for the energy consumption prediction, are made adaptive in order to improve the robustness. The algorithm is validated on a large scale experiment, using two different commercially available electrical vehicles. During the experiment over 4000 kilometer is driven by different drivers. The results are showing a significant improvement (70% error from OEM versus 5% error from TNO) of the predicted range, compared to the original range prediction.