The estimation of the coefficient of performance (COP) and flow ratio (FR) of absorption cooling systems is an important and complex process. In order to simplify this complex process, a gene expression programming (GEP) model was used. In the present study, GEP has been utilized to evaluate the COP and FR of the LiBr-H2O absorption cooling system. The data used as input parameters consisted of evaporator temperature, condenser temperature, absorber temperature, generator temperature and strong and weak solution concentrations. According to these input parameters, in the GEP models, the COP and FR of the LiBr-H2O absorption cooling system were predicted. The training and testing results in the GEP model have shown an acceptable potential for estimating performance of the absorption cooling system in the considered range.