The relationship of silicon epitaxial growth, following decomposition in hydrogen, to classical heterogeneous reaction theory is discussed. The expected behavior of mass transfer control which shifts to a surface effect rate control is found to fit the observations. No conclusion is made whether the surface rate control is, say, due to reaction rate or adsorption rate, but the results appear to favor the adsorption rate of . Experimentally a competing etching reaction is found which appears to explain and production. The diffusion of a reactant away from the surface appears to play the role of reducing the deposition reaction rate, and at high halide concentrations may be responsible for the etching reaction predominating. The variation of the surface morphology with substrate temperature appears to be explainable from the kinetics. The low temperature morphologies of etched and deposited slices were found to be similar.