The liquid desiccant dehumidification system (LDDS) is promising by removing extra air moisture effectively with absorption. By significantly affecting the film wetted area, the desiccant contact angle of the liquid/air interface on the working plate is critical for the system performance. Its value accuracy also seriously impacts the prediction accuracy of theoretical models. However, previous research on the liquid desiccant contact angle was limited, and the results of current predictions were questionable. This paper tested the contact angle of commonly used desiccants, Lithium Chloride (LiCI) and Lithium Bromide (LiBr) aqueous solution, on the stainless steel plate, and developed empirical formulas for predicting the angle. For both desiccants, the contact angle increased with the solution mass concentration and decreased with temperature. With the increase of plate surface roughness, the angle first increased and then decreased. These parameters showed much obvious effects on the angle of LiCI. For the effect of roughness, the maximum changing rate of LiCI was 0.13 degrees/nm compared with 0.07 degrees/nm of LiBr. Errors between new formulas and experiment ones were acceptable, showing an average absolute error of 12.3% for LiCI and 9.4% for LiBr. Furthermore, the effects of contact angle on the film wetted area were investigated experimentally with a single channel regenerator. When the plate surface was hydrophilic, the initial film width was significantly enlarged, with a rate of -2.0 x 10(-3) m per degree, and film contraction along the flow direction was restrained due to the small contact angle. So, the film wetted area increased with an average changing rate of -9.7 x 10(-4) m(2) per degree, and the mass transfer was effectively enhanced by 2-3 times. Effective measures to reduce the contact angle were also proposed. (C) 2015 Elsevier Ltd. All rights reserved.