This paper motivates and applies a variant of the Bayesian Bootstrap Multivariate Regression by Heckelei and Mittelhammer (2003) to a Japanese meat demand specification with endogenous regressors. The methodology is first given an alternative and more elegants motivation and then extended to incorporate microtheoretic restrictions and to apply in the context of a simultaneous equation models. The results of the application are compared to results based on an earlier approach used by Heckelei, Mittelhammer, and Wahl (1996).