Bayesian methods for specification analysis or diagnostic checking of the simultaneous equation model are formulated and applied in analysis of two models. In this work, a direct Monte Carlo simulation approach is employed to compute exact posterior distributions of parameters measuring discrepancies from specifying assumptions, e.g., identifying restrictions exogeneity, etc. Also, ,a new approach for calculating the posterior distributions of a structural equation's parameters is developed and applied using Monte Carlo numerical methods. It is concluded that the methods developed will permit convenient computation of exact finite sample specification error and estimation results for simultaneous equation models.