Effective identification of the source of process variation for a multivariate process is an important research issue and has attracted considerable attention in recent years. Several studies have been devoted to the identification of the source of the process variation, dealing with multiple quantitative data; however, little work has been carried out that deals with multiple correlated counts data. Unlike most of the current research studies, which focus on the identification of the source of process shifts for a multivariate normal process, this study pays attention to identifying the source of process shifts for a multinomial process. A simple detection procedure that utilizes the M test method is proposed for identifying the source of process variation for a multinomial process. An illustrative example is provided to show how to apply the proposed method in practice. Experimental simulation results reveal that the proposed approach is able to effectively identify the source of proportion shifts for a multinomial process.