Self-knowledge is a topic of perennial interest, with roots in ancient traditions that stressed the value of knowing oneself. In contemporary psychological science, self-knowledge continues to play a central role across various domains, including interpersonal relationships, moral behavior, and health. Despite its importance, many fundamental questions remain. We conducted a preregistered, expert-based consensus process to address four key gaps in research on self-knowledge: its (1) conceptualization, (2) measurement, (3) outcomes, and (4) changeability. Seventeen experts from diverse subfields of psychology participated in a structured Delphi process, guided by four facilitators and an external advisor. The process unfolded in four stages: an initial survey of open-ended responses, a survey rating preliminary themes, a three-day in-person meeting, and a final survey assessing revised themes. Across these stages, the panel developed a consensual definition of self-knowledge as the extent to which a person has accurate perceptions of their own relatively stable characteristics and momentary states (e.g., personality features, abilities, affect, motivations, preferences, physical appearance, behaviors). Experts further agreed that self-knowledge is largely domain-specific, context-dependent in its benefits, and malleable in principle though difficult to change in practice. Measurement was identified as a central challenge, with experts outlining several avenues for refinement in future work. Consensus was weaker regarding the existence of a domain-general factor of self-knowledge and shared underlying processes across domains. Overall, the findings clarify where experts converge, where debates persist, and what should be prioritized in future research, providing a crucial foundation for advancing the study of self-knowledge across fields.