The capillary pressure is the key parameter to affect the inversion accuracy of the water-oil relative permeability curve. The existing analytical inversion methods have neglected the influence of capillary pressure, which may cause low precision for the estimated relative permeability curve in some cases. On the basis of the numerical inversion method for the water-oil relative permeability curve established in part 1 (10.1021/ef300018w), taking the one-dimensional radial numerical experiment for example, the rules of relative permeability variation and influence of different displacement conditions on relative permeability deviation when neglecting the capillary pressure are investigated. With regard to water-wet cases whose oil-water viscosity ratio is greater than 1.5, it indicates that the estimated water-phase relative permeability curve is higher and the estimated oil-phase relative permeability curve is lower compared to the true relative permeability curve when the capillary pressure is neglected. The main displacement conditions influencing the inversion accuracy of the relative permeability curve include the injection rate, average permeability, and shape factor of the core sample. As the injection rate increases, the degree of relative permeability deviation caused by neglecting the capillary pressure becomes smaller. Moreover, the deviation trends of the water-oil relative permeability curve are the same as those of the increasing injection rate when average permeability decreases or the shape factor of the core sample increases. Finally, the orthogonal experimental design technique is used to establish the experimental conditions considering the combined effect of multiple factors, and then the water-oil relative permeability curve under every experimental condition is estimated implicitly. On this basis, the multivariate analysis is performed to obtain the threshold value charts of radial displacement experimental parameters, such as the injection rate, average permeability, and shape factor of the core sample, and their corresponding rational value domains are also achieved, which can be used to reduce the influence of neglecting capillary pressure data as much as possible and provide a calculation theory for estimation of the water-oil relative permeability curve accurately.