In systems where safe and effective operation is critical, human error is often cited as a source of failures. The modern perspective on human error is that their associated failures are caused by larger systemic and design problems, not the fault of the person. Within this perspective, classifying human error has utility because it helps analysts identify how and why errors occur so that they can be appropriately addressed in system design. This chapter discusses the science of human error classification. This includes a discussion of leading phenomenological, genotypical, and task-contextualized human error taxonomies. It also describes how concepts from these taxonomies can be used as integral parts of human reliability analysis and accident reporting, two common applications of human error taxonomies. The chapter further discusses emerging classification systems that seek to understand the failures in the automation of complex systems and their relationship to induced human error. Finally, the chapter discusses potential future directions for human error classification.