This article presents a joint multivariate model of the life-cycle interactions between health and socioeconomic status and addresses concerns that emerge in the analysis. A key complicating factor is that outcome variables in the data are discrete, and hence the likelihood function is analytically intractable. Another problem is the presence of endogeneity and simultaneity, which, in conjunction with the discrete dependent variables, renders standard estimators inapplicable. Motivated by these difficulties, we present a simulation-based estimation method that circumvents the intractability of the likelihood, makes inference possible, and allows joint modeling and simultaneous estimation of the interactions of interest. These aspects of the inferential framework are of fundamental importance in dealing with misspecification issues. The methodology is also appealing because it is flexible, computationally efficient, and can be easily modified to capture a variety of interactions in the system.