Understanding the three-dimensional structure (3-D) of GPCRs (G protein coupled receptors) can aid in the design of applicable compounds for the treatment of several human disorders. To this end, several 3-D models have been obtained in recent years. In this work, we have built the rat adenosine receptor model (rA_2AR) by employing computational tools. First, the 3-D rA_2AR model was built by homology modeling using the human adenosine receptor (hA_2AR) structure (PDB codes: 3EML) as a template. Then, the rA_2AR model was refined by molecular dynamics simulations, in which the initial and refined 3-D structures were used for molecular docking simulations and Quantitative structure-activity relationship (QSAR) studies using a set of known experimentally tested ligands to validate this rA2AR model. The results showed that the hindrance effect caused by ribose attached to agonists play an important role in activating the receptor via formation of several hydrogen bonds. In contrast, the lack of this moiety allows blocking of the receptor. The theoretical affinity estimation shows good correlation with reported experimental data. Therefore, this work represents a good example for getting reliable GPCR models under computational procedures.