The annual performance factor (APF) is a new performance index of heat pump type air conditioners (HPACs) based on multiple testing conditions which may result in orders of magnitude larger workload in the product development for enhancing the APF to be close to the maximal value. The development process can be accelerated if the maximum APF can be predicted. The purpose of this study is to present a fast prediction method with a high precision for the maximum APF of HPACs based on a single group of experimental data. The idea of achieving fast prediction is to establish a simplified lumped parameter model for the calculation of capacities and efficiencies of HPACs, and to present a simplified approximate APF correlation by eliminating all the intermediate parameters from the complicated APF equations defined in standards of HPACs. A group of experimental data is applied in the prediction model to improve the precision of the fast prediction method. Application of this method for a typical HPAC shows that the predicted maximum APF of the HPAC agrees well with the tested maximum APF, the maximum deviation of predicted capacities and efficiencies from the experimental data is within ±1.4%; and the duration of prediction for maximum APF is within 12 s on a typical personal computer, significantly saving the time of HPAC development.