A methodology for estimating a state's gross state product (GSP) from primary data sources is presented and applied to Ohio. The social accounts for estimating GSP are used as a skeletal framework orf building a 27 equation econometric model of Ohio. The model is essentially an interdependent system consisting of behavioral equations in the following sectors: consumer, investment, state fiscal, output and personal income, and federal income tax. The sample period for most of the equations is 1949–1963. The adequacy of the model is assessed by the traditional tests of significance of the structural coefficients. In addition, the reduced form of the model is used to estimate the values of the endogenous variables for the sample period. Finally policy analysis is performed by considering alternative assumptions on one exogenous variable, prime military contracts awarded in Ohio, and determining its impact on each of the 27 endogenous variables.