I will describe what is the Matrix Element Method (MEM) and how it can be used for LHC searches and measurements. The MEM allows the direct evaluation of likelihoods for a collision event to arise from different underlying QFT hypotheses, thus allowing in principle maximal use of detector information. The goal is to discuss how the MEM is used in Run II searches in CMS, how it compares to other multivariate approaches and what possibilities there are for extending this method to NLO.