The number of photovoltaic (PV) generators has been significantly increased worldwide as a main renewable energy resource. The output power of PV generators mainly depends on the solar irradiance, module temperature, and field conditions. Therefore, the maintenance of the PV system is necessary to improve the reliability in its use. In this paper, a novel method to diagnose output power lowering in a PV array is proposed to keep the good performance of its output power based on the relationships between PV module temperature, current, and voltage under maximum power point conditions. In particular, the proposed method does not require the accurate information in specification on PV module and sensors for measuring irradiation levels, which are required in conventional monitoring systems. This results in the reduction of cost. Then, it is effectively applied to a practical PV system by using the Kalman-filter algorithm. The performances of the proposed diagnosis method are evaluated and verified with the laboratory experiment and field tests as well as the PSCAD/EMTDC-based simulation test.