All criteria used to diagnose a psychiatric disorder should contribute to distinguishing cases from noncases. The principal of parsimony argues for defining a disorder with as few criteria as possible. Thus, criteria that do not contribute to the case-noncase distinction should be eliminated from the list of defining features because they unnecessarily increase the complexity of the definition of the disorder. In polythetically defined disorders such as major depressive disorder (MDD), diagnosis is based on the presence of a minimum number of features from a list. For a criterion to be retained on such a list, it should contribute to distinguishing between individuals with and without MDD. Simply demonstrating that a criterion is significantly more common in individuals with MDD than individuals without MDD is not a sufficient demonstration of its necessity. Rather, to demonstrate an impact on diagnosis, it should be shown that eliminating the criterion from the list results in individuals being reclassified from a case to a noncase. A criterion does not contribute to determining caseness if its elimination does not result in diagnostic reclassification. The goal of this report from the Rhode Island Hospital Methods to Improve Diagnostic Assessment and Services project was to determine if any of the criteria of MDD are candidates for elimination because of their lack of impact on diagnosis. The results indicated that the symptoms of increased weight, decreased weight, and indecisiveness rarely influenced diagnostic classification and thus are candidates for elimination.