The object of this work was to develop a linear two-point reactor model from an existing nonlinear model and determine optimum parameters for both models in comparison to actual plant data. The importance of keeping the nonlinearity to accurately describe the xenon oscillation problem was also examined. The best chi-square fit to the actual test data was established as the guiding criterion. The nonlinear model initially produced an excellent chi-square fit of 0.708 against plant data. This value could only be reduced through parameter variation to 0.695, approximately 2%. The linear model resulted in a diverging oscillation at the initial one group estimates. Linearizing the oscillatory behavior of the xenon burnup coefficient was attributed to the decrease in the stability of the model. The major flaw of the linear model was the dependence of the nonlinear model for initial conditions. / M. S.