Amniotic fluid embolism remains one of the most devastating conditions in obstetric practice with an incidence of approximately 1 in 40,000 deliveries and a reported mortality rate ranging from 20% to 60%. The pathophysiology appears to involve an abnormal maternal response to fetal tissue exposure associated with breaches of the maternal–fetal physiologic barrier during parturition. This response and its subsequent injury appear to involve activation of proinflammatory mediators similar to that seen with the classic systemic inflammatory response syndrome. Progress in our understanding of this syndrome continues to be hampered by a lack of universally acknowledged diagnostic criteria, the clinical similarities of this condition to other types of acute critical maternal illness, and the presence of a broad spectrum of disease severity. Clinical series based on population or administrative databases that do not include individual chart review by individuals with expertise in critical care obstetrics are likely to both overestimate the incidence and underestimate the mortality of this condition by the inclusion of women who did not have amniotic fluid embolism. Data regarding the presence of risk factors for amniotic fluid embolism are inconsistent and contradictory; at present, no putative risk factor has been identified that would justify modification of standard obstetric practice to reduce the risk of this condition. Maternal treatment is primarily supportive, whereas prompt delivery of the mother who has sustained cardiopulmonary arrest is critical for improved newborn outcome.