This paper studies the efficient output feedback model predictive control (OFMPC) via adaptive event-triggered mechanism for NCS with the data dropout and bounded disturbance. First, an adaptive event-triggered mechanism which utilized an adaptive threshold is considered together with the data dropout and bounded disturbance in NCS. Second, the estimator gain of state estimator is provided by solving a linear matrix inequality (LMI) which guarantees the stability of estimator error. Moreover, an off-line optimization problem is established to get the upper bound of estimator error. Then, through two off-line and one on-line optimization problems, the corresponding the efficient MPC algorithm, which increases the initial feasible set and reduce computation, is obtained. Finally, the provided algorithm about NCS is validated by simulation experiments.