The concern of this paper is with the choice of optimal item selection methods for criterion-referenced tests. Three classes of methods are examined. It is first shown how methods based on classical parameters, such as item difficulty and item-test correlation, have been used for this purpose, and some criticism of these methods is formulated. Next, methods based on item characteristics curve theory are elucidated. These methods have important advantages inasmuch as they are well-suited to analyze the local properties of items at the mastery score. Finally, attention is called to the fact that selecting items for mastery decisions should also allow for the utility of the decision outcomes and the distribution of the true scores. It is shown how this can be taken into consideration by approaching the item selection from a decision-theoretic point of view.