The main object of this paper is to introduce a method to identify outliers in multivariate data set. The robust method which is reviewed in this paper is Minimum Covariance Determinant (MCD). Then, two important characteristics of the robust estimators, i.e. the breakdown point and the influence function, are defined. We also present the consistency and finite–sample correction factors for the MCD estimator. Finally, it is shown that MCD method is better than a classical method for detecting outliers of a real data set.