Self-assessment has been held out as an important mechanism for lifelong learning and self-improvement for health care professionals. However, there is growing concern that individual learners often interpret the results inaccurately. This idea has led to skepticism that self-assessment in its current form can ever be truly useful for lifelong professional development. We examine the proposal that self-assessment can and should be made more effective. First, relevance should be improved. The process should be tied more explicitly to the individual's actual practice profile, rather than being loosely relevant to broader constructs around the permitted scope of practice (eg, certification or licensure). In addition, self-assessment should include not only knowledge and reasoning but also what is done every day in practice, thereby broadening from competence in simulated settings to performance in real settings. Second, the impact of self-assessment should be substantially strengthened by periodic external validation of self-assessment results, together with goals set as a result and plans for further improvement. This offers to the individual the very tangible benefit of satisfying external mandates (eg, licensure and certification). In addition, impact should be reinforced by linking the results of self-assessment to subsequent learning activities including Continuing Medical Education (CME). Although these enhancements individually may not cure all of what ails self-assessment, they might ensure greater effectiveness for the purposes of lifelong learning.