To promote recovery in larger numbers of persons with schizophrenia, considerable research will be required in generating and validating a variety of operational definitions of the construct of 'recovery'. Definitions will vary both in terms of the inclusiveness of their dimensions and their use of dimensional or categorical criteria.When categorical criteria are used, differences in quantitative thresholds for meeting the criteria for recovery will require long-term follow-up studies to choose from. A multi-modal approach to validation of criteria for recovery will likely be most desirable.There will be the need for research and development on instruments for measuring psychosocial functioning, including surveys and focus groups that can clarify the range of normative behavior in the general population using these instruments and their psychometric properties. Studies for identifying those who recover from schizophrenia will need to be longitudinal and prospective, including controlled clinical trials, effectiveness and field studies that view recovery as unfolding over many years in dynamic ways. One benefit of research on the construct of recovery will be to improve the differentiation of schizophrenia as a heterogeneous set of disorders rather than a unitary disease.The role of various malleable factors related to recovery, including treatment and rehabilitation, should lead to advances in services for persons with schizophrenia as well as in training of mental health professionals.