In this paper, the performance of a massive MIMO system from the system security perspective is examined. The telecommunication channel is also considered to be millimeter wave. In the system model, an eavesdropper tries to disrupt information which is communicated between the base station (BS) and users. Due to the presence of an eavesdropper, secure energy efficiency (EE) of the system is taken into consideration. We assume a same channel condition for users and the eavesdropper. First, the secure data rate is calculated, which is obtained from the difference between the user’s rate and the eavesdropper’s rate and has a positive value. The optimization problem in this paper is related to maximizing the secure EE by using appropriate power allocation to users, for which two constraints are defined, which include maximum transmission power and minimum user’s rate. The problem in question is a non-convex and constrained problem; hence, it is difficult to solve. Using nonlinear programming techniques and the Lagrange dual function method, the optimization problem becomes convex and unconstrained. Finally, an iterative algorithm is proposed as a numerical solution of the optimization problem to confirm authors’ claim that the desinged method works well with a high EE performance.