Abstract-In this paper, we develop an uplink pilot and downlink link adaptation approach to improve the energy efficiency (EE) of mobile users in time division duplexing (TDD) multiuser multiple input and multiple output (MU-MIMO) systems. Assuming reciprocity between uplink and downlink channels, the downlink transmission is based on uplink channel estimation. While more uplink pilot power ensures more accurate channel estimation and better downlink performance, it incurs higher energy consumption of mobile users. This paper reveals the relationship and tradeoff among pilot power, channel estimation, and downlink link adaptation that achieves the highest energy efficiency for mobile users. We show that the energy efficiency of different users can be decoupled because the downlink average throughput of each user is independent of the pilot powers of other users and energy-efficient design can be done on a peruser basis. Based on the analysis, we propose an uplink pilot and downlink link adaptation algorithm to improve the EE of mobile users. Simulation results are finally provided to demonstrate the significant gain in energy efficiency for mobile users.