Application of a θ-difference technique to the finite-difference solution of xenon-induced spatial transients has been made, which shows a substantial improvement in the accuracy of the calculated stability index and oscillation period. Virtually no correction is necessary for time-step lengths up to two hours, so an accurate simulation of experimental tests can be performed explicitly in the time domain with fairly crude time-step lengths. A simple expression was obtained for the optimum value of the parameter, θ, that can minimize the calculational error for a broad range of the core stability. The method is expected to be applicable for controlled xenon transients as well as for free-running oscillations.