An approach to simultaneous optimization of assignments of subjects to treatments followed by an end-of-mastery test is presented using the framework of Bayesian decision theory. Focus is on demonstrating how rules for the simultaneous optimization of sequences of decisions can be found. The main advantages of the simultaneous approach, compared to the separate approach, are the more efficient use of data and the fact that more realistic utility structures can be used. The utility structure dealt with in this combined decision problem is a linear utility function. Decision rules are derived for quota-free as well as quota-restricted assignment situations when several culturally biased subpopulations of subjects are to be distinguished. The procedures are demonstrated with an empirical example of instructional decision making in an individualized study system that involves combining two elementary decisions. A 34-item list of references, three data tables, and one figure are included. (Author/TJH)