Foamy oil flow that occurs in heavy oil reservoirs under solution gas drive in the primary production process, shows an anomalous production performance compared with that in conventional solution gas drive reservoirs [Maini, 2001]. Although several laboratory experimental studies have been carried out to research the mechanisms of foamy oil flow in physical models, and the production performance effects of pressure depletion rates on foamy oil flow have been studied, it remains difficult to understand the process clearly. Meanwhile, many oil samples used in previous studies were on heavy oil-methane systems or heavy oil-CO2 systems, rarely on heavy oil-propane systems. Therefore, it is of crucial importance to understand the solution gas drive mechanisms in heavy oil reservoirs with different hydrocarbon solvent gases (methane, propane, and mixture) under different constant pressure depletion rates, so as to guide a better cold heavy oil production. In this study, pressure depletion tests on foamy oil flow were conducted in two kinds of apparatus, namely the Pressure-Volume-Temperature (PVT) system and the sandpack system, to investigate the effects of the pressure depletion rate on different heavy oil-solvent systems in bulk phase and porous media, respectively. Pure methane, pure propane, and a mixture of methane and propane were recombined into a Manatokan dead heavy oil sample to generate live oil samples, respectively. For the heavy oil-pure solvent system, both the PVT tests using the Constant Component Expansion (CCE) approach under constant volume depletion rates and the sandpack tests under constant pressure depletion rates were conducted to examine the effects of different pressure operation schemes on improving foamy oil recovery efficiency. For the heavy oil-mixture system, only sandpack tests were developed under constant pressure depletion rates to study the mixture effects on foamy oil flow. For each heavy oil-solvent system, four different volume or pressure depletion rates were undertaken; in total, eight PVT tests and twelve sandpack tests were carried out in this study. The experimental results showed that for the PVT tests, the plots of pressure versus time elapses and volume changes versus pressure declines of the heavy oil-methane system were much smooth than those of the heavy oil-propane system, which means the volume increase rates of the oil mixture and bulk phase are synchronous in the heavy oil-methane system. However, in the heavy oil-propane system, the volume increase rate of the oil mixture was lower than that of the volume increase rate in the bulk phase. Due to the high solubility of propane in heavy oil, the nucleated bubbles were trapped in the heavy oil and it was difficult for them to evolve out; the propane was recombined into the live oil simultaneously with undergoing the dynamic process undergoing, resulting in appearance of fluctuations in the plots. For the sandpack tests, the trend of oil production recovery factor is different from former researches which indicated that the oil recovery factor has a proportional relationship with pressure depletion rates. In this study, the oil recovery factor plot has a non-linear relation with pressure depletion rates, and there are summits in the recovery plots for the three heavy oil-solvent systems. Finally, for the three heavy oil-solvent systems a brief chart, from which pressure depletion rates can be optimized for the foamy oil flow and Cyclic Solvent Injection (CSI) processes, was developed.