Minimum variance quadratic unbiased estimators (MIVQUE's) of variance components from unbalanced data are obtained for the one-way classification random model under normality. Explicit, computable expressions are given for the estimators, their variances, and their covariance. The variance expressions provide readily-calculated lower bounds for the variances of any quadratic unbiased estimators of the variance components. For unbalanced data, the estimators are functions of the data and of constants σao2 and σe2, taken as a priori estimates of the variance components σa2 and σe2. The estimators are, for unbalanced data, only locally minimum variance, i.e., they are only minimum variance when σao2 = σa2 and σeo2 = σe2. However, numerical results suggest that the “MIVQUE” of σa2 may have much smaller variance than the usual ANOVA estimator with unbalanced data, even when σao2 and σeo2 deviate considerably from σa2 and σe2 respectively. In contrast, the ANOVA estimator of σe2 seems to have smaller variance than the “MIVQUE” unless σao2 and σeo2 are choices close to σa2 and σe2.