Online channels generate frictions when selling products with non-digital attributes, such as apparel. Customers may be reluctant to purchase products they have not been able to try on and those customers who do purchase may return products when they do not fit as expected. Virtual fitting-room technologies provide information about how a product fits a particular customer and promise to mitigate some of the frictions the information gap generates in the retailers' supply chains. By implementing a series of randomized field experiments, we study the value of virtual fit information in online retail. In our experiments, customers are randomly assigned to a treatment condition where virtual fit information is available or to a control condition where virtual fit information is not available. Our results show that offering virtual fit information increases conversion rates and order value, and reduces fulfillment costs arising from returns and home try-on behavior, i.e., customers ordering multiple sizes of the same product. We explore mechanisms through which providing virtual fit information helps customers and retailers. We argue that the virtual fitting tool creates spillovers even to products that are not available for virtual try-on, increases loyalty, helps customers better parse their choice sets, and reduces uncertainty by providing size recommendation.