Desiccant dehumidification has been given increasing interest in the air conditioning industry. Compared with conventional vapor compression air conditioning systems, desiccant dehumidification saves energy by separating humidity control from temperature control and also improves the indoor air quality as a good filter. This research explores the potential of applying desiccant dehumidification systems in buildings with less energy consumption. As the first step, the adsorption mechanism is explored and desiccant material properties are obtained based on a literature review. The heat and mass transfer in the desiccant - moist air system is well understood and modeled using both pseudo-gas-side controlled (PGC) transfer coefficients and semi-infinite transfer coefficients. Compared with experimental data, the model well predicts single processes while the prediction for cyclic processes is acceptable for practical applications. This model provides a useful tool for two purposes: analysis of desiccant unit 's performances and optimization of the design and operations of a unit. Based on the semi-infinite body theory, the semi-infinite model provides a way to simplify the solid-side diffusion resistance. A temperature control strategy is proposed to improve the mass transfer efficiency. A design in which the desiccant temperature is controlled in sections is tested using the model developed before. Simulations show that temperature control enhances mass transfer. Using the model, parametric analysis is conducted on a temperature-control led packed-bed desiccant unit. The effects on dehumidification performances of processing air mass flow rate, regeneration temperature and cycle time are studied. Parametric analysis gains insight into the correlations and interactions between different operation parameters.