Abstract : In the study of inventory policies an interest is shown not only in the mean values of such important random variables as number of shortages per time period, but also in their variance and covariance properties. Such additional properties are of interest in interpreting the stability of an expected value, under assumed inventory policies and parameters, and in using stochastic or Monte Carlo models to calculate estimates of the expected values by sampling techniques. In this paper, comparatively simple inventory models are examined and the expected value, variance, and selected covariance and correlations of the random variables representing stock on hand, shortages per period, overages per period and reorder quantity are derived.