Three experiments investigated the relation between recognition of specific cases and categorization in a double-task paradigm that requires both types of information (Estes, 1986b). Results indicated that recognition and categorization were often affected differently by experimental variables. However, mental models used in categorization sometimes hindered development of experiential (case-based) knowledge, leading to lower levels of case recognition and suboptimal categorization performance. When mental models were complex or difficult to discover (non-salient), subjects often used experiential knowledge to classify into categories, resulting in dependence between categorization and recognition. A model of interactions between the two tasks is proposed that postulates two separate but interacting types of knowledge.