The paper develops a dynamic general-equilibrium model of Schumpeterian growth which is fueled by industrial and agricultural R&D. The former is private and results in better production processes, whereas the latter is government-financed (public and applied) R&D and generates better crop varieties. The arrival of innovations in either sector is stochastic. The model is used to calculate the steady-state equilibrium and the growth-maximizing mix of agricultural and industrial R&D investments. It is also used to highlight the properties of social rates of return (ROR) of R&D which are based on partial-equilibrium calculations. These measures overestimate the true agricultural social ROR and underestimate the true industrial social ROR of R&D investments. This systematic bias increases with the size of the R&D project under evaluation.