Several approaches to the formulation and estimation of dynamic factor demand systems under non-static expectations on the exogenous variables in the firm’s decision process have been suggested. Among those approaches there are trade-offs in terms of statistical and computational efficiency, the generality with which the technology and the expectation formation process can be specified, and in terms of informational requirements. This paper analyzes the trade-offs among three alternative approaches in terms of their statistical and computational efficiency within the context of a Monte Carlo experiment.