Diagnostic radiology is a medical specialty that is primarily devoted to the diagnostic process, centered on the interpretation of medical images. This book reviews the high level of uncertainty inherent to radiological interpretation and the overlap that exists between the uncertainty of the process and what might be considered “error.” There is also a great deal of variability inherent in the physical and technological aspects of the imaging process itself. The information in diagnostic images is subtly encoded, with a broad range of “normal” that usually overlaps the even broader range of “abnormal.” Image interpretation thus blends technology, medical science, and human intuition. To develop their skillset, radiologists train intensively for years, and most develop a remarkable level of expertise. But radiology itself remains a fallible human endeavor, one involving complex neurophysiological and cognitive processes employed under a range of conditions and generally performed under time pressure. This book highlights the human experience of error. A taxonomy of error is presented, along with a theoretical classification of error types based on the underlying causes and an extensive discussion of potential error-reduction strategies. The relevant perceptual science, cognitive science, and imaging science are reviewed. A chapter addresses the issue of accountability for error, including peer review, regulatory oversight/accreditation, and malpractice litigation. The potential impact of artificial intelligence, including the use of machine learning and deep-learning algorithms, to reduce human error and improve radiologists’ efficiency is also explored.