This chapter makes recommendations towards an approach for comparing human and artificial intelligence (AI) capabilities. It stresses the need to compare domain knowledge and skills rather than broad, higher-order abilities such as intelligence. The chapter provides the theoretical and empirical foundations of domain knowledge and skills assessments. It discusses methodological challenges arising from humans’ use of tools, differences in learning between humans and AI, and the inaccuracy of skills assessments at high performance levels. Finally, the chapter proposes an assessment strategy that draws on tests developed for jobs subject to licensing examination.