Many new products (e.g., PDA phones) share features with multiple categories, but are also significantly different from each of these categories. When consumers encounter such a product, they may create a new subcategory (e.g., smart phones) to accommodate it. In such situations, consumers must decide where to position the new subcategory. We develop the Category Activation Model (CAM) to predict where within a category structure consumers are likely to position a subcategory that they have created to accommodate a new product that could potentially belong to multiple product categories. We report two studies that empirically test predictions derived from the CAM. In the first study, we experimentally test the prediction that the new subcategory is likely to be positioned under the relevant category that already contains the greatest number of subcategories. Furthermore, we show that this effect occurs because categories that contain more subcategories are relatively more accessible. In the second study, we analyze category structures that lab participants generated for 100 existing products, and show that, consistent with the CAM's predictions, participants were likely to position these new subcategories under categories that already contained many subcategories. We discuss theoretical and managerial implications of our findings.