. are among the pioneers in commercial exploration of Canadian Athabasca oil sands using the Clark Hot Water Extraction (CHWE) process (Bowman, 1969; Hepler and Smith, 1994). In this process, mined oil sands are digested in hot water under mechanical agitation in a tumbler. Small quantities of caustic are added to the slurry to facilitate bitumen detachment (liberation) from sand surfaces. Since bitumen and water have similar densities, effective recovery of the liberated bitumen from slurry is accomplished by flotation using air bubbles as carriers. In an aerated slurry, bitumen droplets attach to and engulf air bubbles, which float to the top of the slurry to form a bitumen-rich froth in a primary separation vessel (PSV). Bitumen droplets remaining in the slurry are further recovered using conventional mechanical flotation cells. It is clear that the performance of bitumen flotation dictates the overall bitumen recovery.The aforementioned commercial bitumen extraction process is found to be adequate for processing estuarine ores, which contain more than 10% bitumen by weight and a small amount of fine clays. This type of ores is often referred to as "good ores". In this case, a bitumen recovery of greater than 90% can be achieved even with little or no caustic addition (Schramm and Smith, 1987; Hepler and Smith, 1994). In contrast, the existing technology becomes inefficient when processing marine ores, which contain less than 8% bitumen and a relatively large amount of fine clays. A much higher caustic dosage greater than 0.05?40 by weight of the oil sands processed is often added for processing marine ores. But the maximum bitumen recovery achievable remains much lower than 90% (Sanford, 1983; Schramm et al., 1985; Bichard, 1987; Schramm and Smith, 1987). The marine ores are therefore referred to as "poor ores" to characterize their poor processibility.Extensive research work has been conducted in an attempt to understand the causes of poor processibility of marine ores in comparison with estuarine ores. Most of the previous investigations focused on the electrokinetic and interfacial tension analysis (Isaacs and Smolek, 1983; Chow, 1983, 1985; Hsi, 1989, Hepler and Smith, 1994). These early studies provided adequate guidelines for some systems. However, the results were rather case-specific and could not account for complex phenomena in the flotation vessels (Hepler and Smith, 1994). To have a better control of bitumen flotation, it is necessary to consider as many process variables as possible in a comprehensive and systematic manner.Bitumen flotation occurs in a complex system, involving interactions among solids, water, bitumen and air bubbles in the presence of different surface active and ionic species. These species were either added inten-