[en] The regeneration rate prediction model representing the regenerator performance in a liquid desiccant system was derived by statistically analyzing the empirical data collected from the real liquid desiccant unit operated under various conditions. In order to propose a simple empirical regenerator model with wide valid range, additional experimental data found in open literature were also considered in the model derivation. The lithium chloride (LiCl) solution was used as a working desiccant solution. Response surface methodology was used to identify operating parameters and their interactions affecting significantly on the regenerator performance in a liquid desiccant system. Consequently, a first-order linear regression equation was derived as a function of the major parameters and interactions, which returns the regeneration rate in various operating conditions. The reliability of the proposed model was confirmed via the analysis of variation (ANOVA). The proposed model agreed well with the experimental data and other existing models. - Highlights: • A model was developed for predicting the performance of a packed-bed regenerator. • The model was derived based on experimental data collected from the real operation and open literature. • The model is expected to have a wider operating range than an existing model.