In this paper, we propose a novel method for plant identification using a multiscale crossing representation of leaf contour and venation. By extracting the combined features in multiple scale, the proposed method is capable of representing features from global to local regions with mirror, scale, translation and rotation invariance. Three leaf datasets including the Swedish Leaf dataset, the Flavia Leaf dataset and the Soybean Cultivar Leaf dataset are adopted in the experiments to evaluate the performance of the proposed method. Comparative experimental results show that the proposed method can achieve consistently higher or similar recognition accuracy than the state-of-the-art methods among these leaf datasets, which may indicate a new solution to the leaf identification problem.