Organizations learn by sharing information. While filtering may exclude potentially valuable information, information overload may prevent the adequate identification of important information. Lee andBrooks [1993] report on the introduction of a document classification and information dissemination system for "soft information." There was an initial concern that "users would use "high" priority categories excessively within the grapeVINE system, thereby reducing the effectiveness of the filtering process." In fact, Lee and Brook found "that people tended to undervalue their contributions and not put a high enough priority on their contribution." A large percent of the information in the grapeVINE system was added at low priority, causing the information not to be shared. A major advantage touted of groups is the potential for pooling unshared information and thereby improving task accomplishment. Stasser found that groups tend to discuss topics that they have in common(shared information) more then their unique knowledge (unshared information) [Stewart, 1992]. Based on social validation theory, Stewart [1992] predicted and found evidence that telling someone he or she is an expert, separately and in front of the other members of a group, increased the proportion of unshared information. This paper explores the affect of expertise on organizational transactive memory with respect to the filtering and sharing of information. First transactive organizational memory is briefly discussed, followed by relevant aspects of social validation theory and a description of the filtering and sharing model embodied in Brook's classification and dissemination system. We then describe the experimental design used to isolate the effects of expertise on the filtering and sharing of information, present results, and discuss their implications.