This study proposes a prediction scheme for qualitative evaluation of noisy dynamic process variables from a set of correlated data. The process models of a noisy power generating plant to, w hich this study directly applies incorporate a priori knowledge of noise contamination in the measured terminal voltage and load frequency. A conceptual simplicity of parameter by a least squares computational algorithm is evident in the analysis. Convergence of the iteration was enhanced by postulation of a filter to enable the additive noise to be reduced to a white noise sequence which yielded unbiased estimates of the measured data.