OBJECTIVE: Many Americans use religious activity to cope with stressful life events. Our goal was to review systematically the recent medical literature to assess the role of religion in health outcomes. DATA SOURCES: We conducted a comprehensive literature search using MEDLINE to identify studies published in the English language between January 1999 and June 2003 describing the effect of religion on health outcomes. The search strategy used the medical subject headings (MeSH) of religion; religion AND medicine; religion OR intercessory prayer; prayer; prayer therapy; religious rites; faith; medicine, traditional; religiosity; religion AND psychology; and religion AND health. STUDY SELECTION: Religious, but not spiritual, interventions were selected for inclusion. Thus, papers describing interventions such as yoga, meditation, acupuncture, and qigong were excluded. Manuscripts describing randomized controlled trials, clinical trials, and partnerships with faith-based organizations were included. DATA EXTRACTION: We found five randomized controlled trials, four clinical trials, and seven faith-based partnerships that describe the impact of religious intervention on health outcomes. Papers were analyzed by four reviewers using a modified Delphi technique to reach consensus. DATA SYNTHESIS: Religious intervention such as intercessory prayer may improve success rates of in vitro fertilization, decrease length of hospital stay and duration of fever in septic patients, increase immune function, improve rheumatoid arthritis, and reduce anxiety. Frequent attendance at religious services likely improves health behaviors. Moreover, prayer may decrease adverse outcomes in patients with cardiac disease. CONCLUSIONS: Religious activity may improve health outcomes.