Abstract High pressure accelerating rate calorimetry (ARC) tests have been performed on three different crude oils at their respective reservoir pressures. The experiments on medium heavy Clair oil and heavy Wolf Lake oil used clean silica sand, incorporating 3% kaolinite, to represent the reservoir matrix, whereas Athabasca Tar Sand was used in its preserved, virgin state. The Clair oil and Wolf Lake oil tests involved high initial water saturations in the reservoir, representative of a post-waterflooded or a post-steam injection state. A second test on Athabasca Tar Sand used a lower oil (bitumen) saturation, but without an adjustment of the original brine saturation. Although the medium Clair oil exhibited high exothermicity throughout the temperature range, the temperature detected for the onset of low temperature oxidation (LTO) was much higher than that for Athabasca Tar Sand. The calculated activation energies also indicate that Athabasca Tar Sand is very reactive in the LTO region. Furthermore, the overall continuity exhibited by the measured exotherms indicates that all three oils are potentially good candidates for in situ combustion as an oil recovery method. Introduction The most important factor governing the selection of any improved oil recovery (IOR) process is, ultimately, the availability of a suitable fluid to inject into the reservoir. For an increasing number of reservoir situations, air injection is gaining increasing attention for both heavy and light oil recovery. The most recent heavy oil projects are in the Cambay Basin, India(1), using conventional ISC, and the first field pilot at Christina Lake, using the advanced THAI (Toe-to-Heel Air Injection) process(2). There have also been significant developments in high pressure air injection into light oil reservoirs, as evidenced by economically viable air injection projects in the Williston Basin, and project feasibility studies for Barrancas, Argentina(3) and Cantun, Mexico(4). There are many economic and technical benefits derived from air injection – excellent displacement efficiency, mobilization of extra oil ahead of the combustion zone, reservoir pressurization, oil swelling from produced CO2, flue gas stripping, injection gas substitution and, for high pressure hot reservoirs, additional enhancing factors such as spontaneous ignition and near miscibility effects. The accelerating rate calorimeter (ARC) was first advocated as a technique for screening reservoir oil candidates for air injection by Yannimaras et al.(5) The technique was developed as a method for studying the reaction kinetics by following the reactions adiabatically, especially at high pressure. Moreover, the method can provide insights to explain the occurrence of low temperature oxidation (LTO) and high temperature oxidation (HTO). To date, investigations using the ARC technique have focused mainly on light oil applications. In this paper, results are presented for heavy crude oils and also Athabasca Tar Sand, to see if the measures of oxidation reactivity obtained are relevant when considering the application of air injection processes using conventional ISC or THAI. Experimental ARC Equipment The main apparatus consists of a pressure shell, inside which the calorimeter ‘bomb’, or test cell, is placed (Figure 1). The cylindrical test cell is surrounded by a top guard heater, as well as side and bottom guard heaters.