An experimental methodology study of five common mounting techniques and procedures and three calculation methods helped determine the specific X-ray techniques to be used in the semi-quantitative estimation of clay minerals in the study of offshore Lake Agassiz sediments. The aqueous pipette-on-g1ass slide technique with calculation by the direct comparison method was chosen because it offered best and most consisitent results, ease and rapidity of obtaining results and best correlation possibilities. Numerous attempts have been made at subdividing the offshore deposits of Lake Agassiz in southern Manitoba and adjacent parts of the United States. The results of detailed clay mineralogical and textural analyses of samples from the Red River basin suggest a three-fold stratigraphic subdivision. This subdivision is also supported by non-clay mineralogy differences within the sediment and by the limited soil engineering data available. Unit 1, the lowermost offshore unit recognized, is defined primarily on the basis of its low montmorillonite:illite ratio, very fine average grain size and low Atterberg limits. Unit 2 is recognized by its high montmorillonite:illite ratio, very fine average grain size, and high Atterberg limits. Unit 3, the uppermost offshore unit recognized, is defined by its low montmorillonite:illite ratio relatively coarse grain size and low Atterberg limits. These three units can be idenified throughout the sedimentary basin in Manitoba and roughly correlate with units in North Dakota. Although various hypotheses can be suggested to explain the observed stratigraphic and lateral changes within the offshore sediments, probably the most reasonable explanation is that of differential settling of the clay particles caused by salinity changes in the lake and size difference of the various minerals.