A comprehensive laboratory study was conducted for the recovery of heavy oil from a linear (1-D) and a three-dimensional (3-D) physical models, packed with 12.4 degrees API gravity heavy oil, 60000 ppm brine and crushed limestone. A total of 19 experiments were conducted to investigate the effects of injecting CH4 or CO2 along with steam by using these laboratory models. A linear model having a 6.6-cm diameter and 100-cm length was used for the first group of experiments. In the: second group, a 3-D model with 30-cm x 30-cm x 7.5-cm dimensions was used to obtain the process parameters. Temperature distributions in both models, steam and gas injection pressures, and fluid productions were recorded during the experiments. Base runs involving steam injection alone facilitated the interpretation of steam-gas injection results and showed that on the whole, steam-gas injection is a far more promising technique. Lower residual oil saturations were achieved in steam-CO2, and steam-CH4 injection runs, compared to the values obtained with steam alone. Among the variables studied, CO2/steam and CH4/steam injection ratios appeared to be significant factors in heavy-oil recovery. The results of 1-D model tests indicated that the optimum gas-steam ratio for maximizing oil recovery was about 9.4 cc/cc for both steam-CO2 and steam-CH4 processes. For 3-D model tests, it was about 8.7 cc/cc far steam-CO2 process. As no maximum in recovery was found for steam-CH4, no optimum gas-steam ratio can be identified. From 1-D model tests, 30.6% more oil for the steam-CO2 case and 18.7% more oil for the steam-CH4, case were recovered than from the steam-only runs. 130% more oil for the steam-CO2 case and 70% more oil for the steam-CH4, test were obtained compared to the steam-only tests from 3-D model experiments.