Introduction The production of large quantities of tailings sludge r in the extraction of bitumen from oil sands by the Clark I shot water process, poses a major environmental problem with regards to its final disposal. This sludge is retained in tailings ponds where it remains fluid, even after several years. Therefore, the design of an efficient method to transform this fluid waste into a solid sediment is generally considered as a major task to be achieved. Its accomplishment rests on a clear understanding of the behavior of tailings sludge, and on the determination of the influence of a number of parameters, such as its composition. The problem of tailings sedimentation is not limited to the oil sands industry only. It is often present in the chemical processing 1ndustries, the metallurgical industries, or simply the treatment of city wastewaters. An important body of research results is, therefore, available concerning the sedimentation of tailings, at large. For instance, Coe and Clevenger (1916) studied the effect of electrolytes on the settling behavior of slime pulp, and identified two types of settling behavior, Baillie and Malmberg (1969) have patented a process using flocculation followed by centrifugation to remove the clays from the hot water streams. It rests on the use of a number of chemicals and on the adjustment of the pH. Hepp and Camp (1970) have patented the use of a "vacuum precoat filtration" technique after flocculation. Stastny (1973,1974) found that the injection of carbon dioxide into tailings sludge samples promoted their settling by reducing their pH. Also, as a general fact, Baillie and Malmberg (1969) and Schutte (1974) have shown that a reduction of the pH caused flocculation. In general, these works have addressed directly the treatment of actual sludges, for example, the oil sands tailings sludge. The composition of actual oil sands tailings sludge, however, is very complex. The studies made so far do not allow the prediction of its behavior when its composition, or more generally the procedure to extract the bitumen r is changed. The authors believe that the oil sands industry needs model systems, where a number of important parameters can be varied easily and their effect assessed immediately. Ultimately, the best treatment methods of tailings sludge can be derived from such models. The data presented here are a contribution to the achievement of this task, with clay- Fe- water ternary systems as the first model. The clay being addressed here is montmorillonite, but similar studies are underway with kaolinite. The major components of actual oil sand tailings sludge are water, sand, silt, clay, bitumen and naphtha. There are also small concentrations of dissolved organics and of various cations and anions. The word sludge covers aqueous suspensions in the tailings ponds, with a solid content which is not well defined, but is usually considered to range from 5% to 20% by mass. By any measure, this is a high solid content by comparison with suspensions to which the D.L.V.O stabilization theory applies quantitatively.