The original Buckley-Leverett theory has been extended and a more detailed formulation of the waterflood behavior in linear horizontal systems is presented. Particular consideration has been given to the evaluation of capillary pressure effects and differential equations permitting an explicit evaluation of these effects have been derived. On the basis of the developed theory it is recognized that the flooding behavior is dependent upon the length of the system and the rate of injection. At the same time it has been determined that systems of different lengths yield the same flooding behavior if the injection rates and/or the fluid viscosities are properly adjusted or "scaled." It has also been found that the sensitivity of the flooding behavior with respect to rate and length decreases as any one of these factors increases in value and that for sufficiently long systems and high rates of water injection the flooding behavior becomes independent of rate and length, or "stabilized." To such stabilized conditions the theory formulated by Buckley and Leverett is applicable. A number of laboratory flooding tests have been made and good agreement has been found between theory and experimental observations. The experimental results are discussed and it is shown that under field conditions the flooding behavior is usually stabilized. As a result of these findings a procedure is indicated for evaluating field performances either on the basis of tests performed with commonly available core samples or by means of calculations using relative permeability data.