Abstract Reported properties for Athabasca oil sands have historically been in error as the result of core expansion due to release and expansion of gas dissolved in the pore liquid. The resulting high porosities have resulted in underestimation of material weights and volumes. In-situ properties differ considerably from laboratory tests on the best available specimens; in-situ compressibilities are much lower, elastic moduli are higher, shear strengths are higher and water permeabilities are much lower than laboratory data indicate. If truly undisturbed specimens are judged to be necessary, several methods are possible; for example, the use of pressure core barrels, or in-situ prefreezing (liquid N2) and air coring. If slightly disturbed specimens are acceptable, oil-free outcrop samples or cored samples from behind outcrops where gas hasdissipated may provide adequate data. The alternatives to high-quality sampling are to use in-situ test devices or to design without benefit of the best data. Introduction Controversy concerning the true in-situ nature of oil sands has continued for many years. In large part, this controversy is the result of inadequate sampling techniques. Interpretation problems have arisen from the application of high-technology laboratory and analytical methods to specimens of very poor quality. Grossly disturbed oil sand specimens have been tested without knowledge of prior sample history and the results have been assumed to represent true in-situ oil sand characteristics. Millions of dolars have been spent in analysis, model testing, laboratory testing and field-scale tests, but the problem of obtaining the necessary undisturbed samples has not been approached systematically. Underground excavation in oil sand is of considerable interest (Roesner and Poppen, 1978; Smith et al., 1978; and others), and novel technologies of production from oil sand are being investigated (Harris and Sobkowicz, 1978; Haston, 1978; Nicholis and Luhning, 1978). Adequate, samples for laboratory testing will increase understanding of the behaviour of oil sands and therefore will aid in the development of the dvanced technology required for economic development of oil sands. Historical Review of Oil Sand Sampling Ells (1914) reported a "specific gravity" of oil sand of 1.75 as being representative of in-situ conditions. This figure, assuming 17% by weight of bitumen and water, implies a porosity of 45% and a pore space liquid saturation of less than 70%. These implied properties are not possible for rounded quartzose sands which have been buried to depths greather than 25m. Nevertheless, subsequent authors have accepted these figures and deduced that sand grains were "floating" in oil (Ball, 1935). Hume (1947) speculated extensively about the nature of oil sands. Several quotes from his paper will serve to demonstrate the changes in knowledge in the past 30 years and the nature of some of the more common early misconceptions:"...large deposit of bituminous sands and bitumen beds..." (p. 298);"... beds of bitumen occur interstratified with thebituminous sand..." (p. 303);"...The bitumen beds are...bitumen concentrations with low sand content... " (p. 314);