This paper presents adsorption and desorption characteristics of two different working pairs—activated carbon–methanol and activated carbon–R134a—determined experimentally. Dubinin–Radushkevich (D–R) equation is used to correlate the adsorption isotherms and to form the pressure, temperature, and concentration diagrams for both the assorted working pairs. The results show that the maximum adsorption capacity of activated carbon–R134a working pair is 1.21 times that of activated carbon–methanol. Temperature and pressure distribution throughout the adsorbent bed and their variation with adsorption time are also predicted. Use of artificial neural network (ANN) is proposed to determine the uptake from measured pressure and temperature. The back propagation algorithm with three different variants, namely, scaled conjugate gradient (SCG), Pola–Ribiere conjugate gradient (CGP), and Levenberg–Marquardt (LM) and logistic sigmoid transfer function are used, so that the best approach could be found out. After training, it is found that LM algorithm with 11 neurons is the most suitable for modeling adsorption refrigeration system. The adsorption and desorption uptake obtained experimentally are compared with the uptake predicted by D–R equation and ANN modeling.