This paper is concerned with the identification and estimation of the parameters in a dynamic simultaneous equations model with stationary disturbances when both the endogenous and exogenous variables are subject to random measurement errors. A frequency domain approach is suggested to fully utilize the information contained in the data. The first part of this paper explores the identification criteria. The second part of this paper suggests estimation methods for such a model. Both full information and limited information estimation methods are studied and their respective gains and losses are evaluated.