Research germane to product recalls and their causes is limited. With recall rates rising in many industries, it is timely and pertinent to comprehensively investigate recalls. The focus of my dissertation is on product recalls and their causes, with the objective of recall understanding and prevention. I study three important phases in the product recall process at multiple organizational levels in the high-risk medical device industry: plant-level causes, recall decision-making, and causes and effects of firm and regulator responsiveness within the recall event. First, I study the relationship between Food and Drug Administration (FDA) plant inspections and future recalls. Using a 7-year panel dataset and recurrent event Cox proportional hazard and propensity score matching models, I find that adverse plant inspection outcomes serve as warning signs for future recalls. I incorporate FDA investigator experience to identify reasons for, and effects of, investigator complacency in repeated plant inspections. Repeated visits to the same site by an inspector increases the recall risk and also reduces the predictability of inspection outcomes as a leading indicator of future recalls. FDA investigator rotation is shown to be an effective solution to compensate for investigator complacency. Second, I explore behavioral factors thainfluence managers' decision to recall. Recall guidance provided by the FDA allows for broad managerial interpretation so it is crucial to study which factors influence managers to choose to recall. Using actual industry managers with recall experience in a controlled experiment, I find that product defects which are undetectable to physician customers pre-use are more likely to lead to a recall than detectable ones. When managers have a deeper understanding about the root cause of a defect, they are also more likely to recall. I also study individual dispositional factors unique to each manager, and surprisingly find that the level of cognitive reflection, as measured by the Cognitive Reflection Test (CRT), is the most important predictor of a recall decision in the experiment. Finally, I study firm and regulator recall responsiveness. Responsiveness is critical in this domain: the longer a faulty medical device remains on the marketplace, the more consumers are at risk. Using an 11-year panel dataset with time-stamps for over 4,000 recalls, and multiple hazard and fixed effects panel models, I find that higher recall severity leads to slower firm and faster FDA responsiveness. However, taking longer to close a recall reduces a firm's future recalls, and this may be attributed to learning mechanisms. FDA response times also reduce future recalls.