Introduction In situ combustion is an enhanced recovery technique which offers many advantages over other processes, such as: a more efficient overall drive mechanism, less energy consumption and less total environmental impact. Although this process has been extensively studied in both the laboratory and the field since itas initially patented in the 1920's, it has failed to attain the acceptance which it rightfully deserves. This is primarily due to a lack of understanding of how combustion actually works. This article will examine some of the reasons why field projects have failed in the past, discuss several parameters which indicate whether or not problems are occurring with the process and offer suggestions on how the correctoxidation mode can be encouraged in a fireflood operation. Although space does not permit an extensive review of the in situ combustion literature, the interested reader is directed to the excellent reviews by Ramey(1), White(2) and Chu(3, 4). The classic articles by Nelson and McNeil(5) and Gate and Ramey(6) are also required reading for anyone contemplating the design of a field project. The In Situ Combustion Process The traditional concept of in situ combustion is that of a propagating high-temperature combustion zone which displaces oil towards the production wells. The injection gas may be normal air, enriched air (usually pure oxygen) or depleted air (normal air blended with recycled product gas). The fuel for the process is normally assumed to be a coke-like substance which is deposited on the mineral matrix by thermal cracking, or pyrolysis reactions. These reactions occur on the leading edge of the combustion region, where the oxygen concentration is assumed to be zero. The process can be modified by injecting water, either simultaneously or in alternate slugs, with the air. Normal-well combustion occurs when the injection rate of the water is such that all of the water passing through the oxidation zone in the vapour state. When the water rate is sufficiently large that liquid water is passing through the oxidation region, the process is termed superwet combustion, In the superwet mode, the reaction zone propagates as a steam bank. Temperatures in the oxidation zone during dry and normal-wet combustion are generally in the range of 400 to 600 ºC, and thus we normally refer to high-temperature combustion rather than oxidation reactions. For a given oil, the peak temperatures during normal-well combustion are often slightly higher than those for dry combustion. Product gas compositions are similar for both of these burning modes. Because these high temperatures are not present during superwet combustion, the region in which oxygen is primarily reacted is designated as an oxidation zone. For a given oil, the temperature within the oxidation zone and the product gas composition are very dependent on the operating pressure. This is due to the fact that the temperature levels are dictated by the thermodynamic properties of steam, with the actual temperature being somewhere between the bubble and dew point for the localized water and non-condensable gas concentrations within the oxidation zone.