This study aimed to develop and test a performance prediction method using virtual refrigeration cycle generation through the air and indoor-side data of a building automation system and heat pumps in an air source heat pump system of a central air-conditioning system, to which the building automation system is applied. The results of coefficient of performance, Δhcomp,ref, Δhcond,ref, and Δhevap,ref of the measured refrigeration cycle according to outside temperature were confirmed. A similar pattern was confirmed by the prediction results through the virtual refrigeration cycle. Reliability evaluation of the performance prediction results of the air source heat pump system was conducted using NMBE and CvRMSE as proposed in ASHRAE guideline 14. NMBE and CvRMSE were confirmed as -6.33% and 15.99% for the coefficient of performance, 0.91% and 11.78% for Δhcond,ref, -5.53% and 12.48% for Δhevap,ref, and 0.14% and 2.85% for Δhcomp,ref, respectively. Thus, the performance of the air heat source heat pump system using a virtual refrigeration cycle can be predicted. Future research will use the performance prediction model of the air heat source heat pump system developed in this study improve the detection and diagnosis of physical and operational failures in heat pump systems.