A discussion of a model for predicting income distribution across states. The basic assumption is a constant industry wage structure and level across areas; therefore, the amount of human capital is a function of the industry mix in a given area. Variables such as discrimination, education and training, labor supply, and property income were not considered to be of significant influence. The model is tested using data from 51 states and the District of Columbia. The results are analyzed using canonical correlation. This indicates that the basic hypothesis serves very well in predicting income distribution across states and that factors other than industry mix are either not important in determining the shape of the income distribution or are correlated with industry mix in the sample. 26 pp.