The study evalutes public investment in on the job training (OJT) and tries to identify systematic determinants of the profitability of this investment. Profiles of age and earnings by occupation are estimated for the South and other regions, and for whites and nonwhites, from data in the one-in-1,000 census sample and are then adjusted to the regional level using wage survey indexes. Concepts of profitability reflect three kinds of assumptions: (1) transfers (government expenditures for allowances to trainees) are not a cost; (2) transfers are a cost; (3) increases in tax revenue are the sole benefit derived from OJT. The methodology of the study is an aggregate approach to evaluating investment by utilizing detailed information describing individuals. (author/ly)