The calculations involved in estimating variance components from unbalanced data (having unequal numbers of observations in the subclasses) are somewhat tedious for the mixed model and usually not computationally feasible for data having many classifications. This paper outlines the analysis for unbalanced data from a 2-way classification with one classification considered fixed, and relates it to the analysis based on the random model. Simplified computing procedures are given which are suitable for a large number of levels in the random classification and a reasonable number of levels in the fixed classification.