The surface vapor pressure of desiccant solutions is the key parameter for modeling the liquid desiccant system. Usually, the simple mixing rule (SMR) model and the nonrandom two-liquid (NRTL) model are employed by people to calculate the surface vapor pressure of mixture desiccant solutions during simulating the performance of a liquid desiccant system. However, there is little literature caring about the calculation accuracy of the two models for surface vapor pressure of mixture desiccant solution under different circumstances. In this paper, an experimental system is built to measure the surface vapor pressure of LiCl-CaCl2 desiccant solutions under various conditions including temperatures ranging from 25 degrees C to 75 degrees C, total mass concentrations ranging from 5% to 40% and three mass mixing ratios of LiCl to CaCl2 (e.g., 1:1, 1:2, 2:1). The two models for surface vapor pressure of mixture desiccant solution (i.e., the SMR and the NRTL) are compared with the test data. It is found that the SMR model has much better accuracy than the NRTL model in forecasting the surface vapor pressure of LiCl-CaCl2 mixture desiccant solution when the solution mass concentration is lower than 10%, and it is the opposite for the case when the solution mass concentration is higher than 30%. Afterwards, the revised equations for the SMR and the NRTL model are proposed according to the test data within the conditions concerned in this study. The calculation accuracy of the two vapor pressure models is proved to be greatly improved after the revisions. (C) 2016 Elsevier Ltd. All rights reserved.