The online purchase of garments is increasing and a method to enable accurate and immediate garment sizing could improve the customer’s experience. A successful online garments shopping system should provide the capacity of dressing a given garment onto various posed human models with fit/ease information. In this paper, we present a method to “copy” the pose of a source human model to a target human model via a skeleton- matching algorithm. The skeleton is generated automatically according to the anthropometric features. The pose difference is compensated by an affine transformation applied to the skin vertices recursively. The final redressing is conducted according to benchmark matching and a penetration recovery procedure followed by a physical-based drape simulation. The fit evaluation is fulfilled through cutting the segmented human model with a serious of planes. The experimental results validate that this method is an effective approach for predicting dressing style with accurate fit/ease information.