Solvent-based nonthermal heavy oil recovery methods are relatively new and still under development. The central idea in such techniques is to rely on solvent dissolution for viscosity reduction instead,of heating. Among the solvent-based methods, VAPEX appears to be the most promising. In thin heavy oil and bitumen reservoirs, where the thermal processes are likely to fail due to excessive heat losses, VAPEX can be more successful. Nonetheless, the economic viability of VAPEX remains uncertain due to lower oil production rates predicted by scale-up of laboratory model results using the transmissibility based scaling criteria. However, such scaling is far from reality, and our previous experimental work [Yazdani and Maini. SPE Reservoir Eng. Eval. 2005, 8, 205-213] has shown that it underestimates the field rates. This paper presents the design of a very large physical model as well as the results of a set of VAPEX experiments carried out in this model. The results are, interestingly, in very good agreement with the trend of the previous experiments conducted by the authors [Yazdani and Maini. SPE Reservoir Eng. Eval. 2005, 8, 205-213] in smaller models. The results from this large model were combined with the previous results to improve the previously reported empirical scale-up correlation.