This paper describes a data-driven injury cost model (ICM) developed to estimate the costs associated with non-fatal consumer product injuries. The modeling effort combines information by diagnosis from the US Consumer Product Safety Commission’s National Electronic Injury Surveillance System (NEISS) and 17 other large data sets. The ICM contains four aggregated cost components: (1) medical costs, (2) work losses, (3) quality of life and pain and suffering costs, and (4) product liability insurance administration and litigation costs. The ICM estimates societal costs, which are broader than costs to any individual group, such as victims, insurers, or product manufacturers. Costs associated with consumer product injuries are estimated to be approximately $500 billion in 1996, accounting for nearly one-third of the total annual injury costs. We examine injury costs in several ways, including by major product category, by sex and age of victims, by body part injured, by injury diagnosis, and by highest level of medical treatment received. We also rank the 10 leading consumer products that account for injury costs overall and within different age groups. Products such as stairs and floors are among the top 10 for all age groups. Other products, however, are more closely tied to injuries at particular stages of life (e.g., infant/toddler, child, young adult, elderly). These cost estimates are useful in assessing which products and types of injuries impose the greatest costs on society and for identifying areas for focused injury prevention efforts.