Alfalfa ( Medicago Sativa ) is the queen of forage plants. Leaves contain two-third of the protein and 75 percent of the total digestible nutrients (TDN) in alfalfa hay. Harvesting the maximum amount of the leaves is a profitable way to produce alfalfa hay. To prevent losses, particularly the leaves, during alfalfa baling under high temperature and low relative humidity conditions, it seems rewetting moved and dried alfalfa rows could increase moisture content, and hence, reduces the losses in alfalfa baling. The objective of this research was to obtain adsorption and desorption characteristics of alfalfa leaves and stems. The adsorption and desorption moisture content of alfalfa leaves and stems were determined using the dynamic gravimetric method. Determination was made within a range of water activity of 0.11 to 0.84. It is necessary to use at least three temperatures for the safety of the estimations based on the experimental results. These experiments were conducted at three different temperatures of 30 ° , 40 ° and 50 ° C. Seven saturated salts used to provide relative humilities. Data from the experiments were analyzed using non-linear regression techniques in Microsoft Excel software. The SAS and MSTATC softwares were used for data analysis. In this study, the effects of relative humidity and temperature on sorption and desorption equilibrium moisture content (EMC) of the alfalfa components (leaves and stems) were investigated. The experiments were conducted using split plot experiment using a completely randomized design with three replications. The mean comparison was performed using least significant difference (LSD). Several theoretical, semi-theoretical and empirical models have been proposed and used by researchers for fitting EMC data. Four mathematical models consisting of Oswin, Halsey, Henderson and Chung-pfost were used to fit the experimental data. All the models were compared according to three statistical parameters; i.e. adjusted R square (R 2 ), standard error of the estimates (SEE) and mean relative error (MRE). For selection of the best of four models, the model with the least SEE and MRE and the highest R 2 were chosen. Results indicated that the effect of relative humidity, temperature and alfalfa components, as well as their interactions on sorption and desorption equilibrium moisture content were significant ( P 0.01). The isotherm curves were sigmoid in shape, which conformed to type II Henderson model was a good model for moisture desorption of stems at 50 ? C. Keywords: Alfalfa, equilibrium moisture content, relative humidity, adsorption, desorption, isotherm curve