High adverse event rates are a signal of potentially low-quality care that imposes burdens on patients, families, and hospitals. In this article, we examine the relationship between the distinct characteristics of teaching hospitals with adverse event rates among pediatric patients, controlling for patient complexity and severity using 2009-2011 Nationwide Inpatient Sample data from the Agency for Healthcare Research and Quality. We hypothesize that adverse event rates increase with the availability of more complex services and technologies (transplantation and pediatric open-heart surgery); increase as experience of providers decreases (July effect); and increase with residents per bed, a measure of both average provider inexperience and congestion. Using multilevel analysis, we find empirical evidence in support of our three hypotheses. We find that in environments where more learning occurs, more mistakes are made. Identifying high-performing hospitals with large residency programs and complex service lines that have made progress in patient safety and then studying how they have done so should become a priority. These findings should then be adapted within other hospitals through publicly funded mechanisms to improve the quality of care for all children.