Results of a comprehensive research effort to investigate and quantify factors which affect the performance of CO/sub 2/ flood processes in reservoir rocks are presented. An apparatus is described in which displacement of oil from pore networks etched in glass by CO/sub 2/ at high pressure can be observed visually. Results of a series of secondary and tertiary, first contact and multiple contact miscible displacements are reported. Equipment for on-line measurements of mixing in one- and two-phase displacements is described. A new device designed to separate oil and water continuously in a chamber of very low dead volume is also described. Results of one- and two-phase displacements which validate the experimental techniques are reported. Results of a series of PVT experiments to study the effects of changing both the amount of dissolved gas in a crude oil and the system temperature are given. Those results suggest strongly that the extrapolated vapor pressure of CO/sub 2/ is a good estimate of the minimum miscibility pressure if the temperature is below about 140/sup 0/F. Modifications and improvements to the design of the continuous multiple contact apparatus are described, as are preliminary tests of the use of an oscillating quartz crystal to measure viscosities. Results of a series of phase composition measurements to study effects of oil composition on component partitioning in CO/sub 2/-crude oil systems are reported.The theory of the effect of CO/sub 2/-crude oil phase behavior on displacements in an ideal porous medium is tested by predicting the performance of slim tube displacements from independently measured phase behavior and fluid property data obtained with the continuous multiple contact experiment. The quantitative agreement between prediction and experiment is good enough to suggest that the essentials of CO/sub 2/-crude oil phase behavior can be modeled with a small number of pseudo-components.