Abstract: The classical Fuzzy C-Means algorithm（FCM） is based on Euclidean distance function，which can only be used to detect spherical structural clusters.When FCM processes some dataset of high dimension，error probability will be increased.Focusing on above two problems，this paper proposes an improved new algorithm called Fuzzy C-Means based on Mahalanobis distance function（FCM-M），and adds a regulating factor of covariance matrix to each class in objective function.Using the advantage of Mahalanobis distance，FCM-M algorithm effectively solves the shortcomings of FCM algorithm.There are efficient methods to solve singular values problem for finding Eigen value and eigenvectors of a symmetric matrix or computing pseudoinvertion involved in finding the Mahalanobis distance.Experimental results of data clustering and image segmentation illustrate its effectiveness and feasibility.