The objective of this study was to describe the mental health effects of high ambient temperatures and heat waves, determine whether heat-related morbidity and mortality are increased among people with known mental disorders and identify knowledge gaps to inform targeting of future research. A systematic review of published literature using MEDLINE, Embase and PsycINFO without geographical or temporal limits was conducted. We included studies that explicitly linked high ambient temperatures and/or heat waves to mental health outcomes. Evidence was graded using a National Institute of Health framework. A total of 35 studies were included in the review, among which 34 were observational studies and one was a case-control study. Six broad mental health outcome categories were identified: suicide and heat; bipolar disorder, mania and depression and heat; schizophrenia and heat; organic mental health outcomes and heat including dementia; alcohol and substance misuse and heat; and multiple mental health outcomes/mental health service usage and heat. The studies included were heterogeneous in terms of study design, population, setting, exposure measures, outcomes and location making meta-analysis unfeasible. Fifteen studies showed an increased suicide risk with heat (relative risk [RR] 1.014-1.37 per 1 °C, P < 0.05; r = 0.10-0.64, P < 0.05). Increased risks of mental health-related admissions and emergency department visits at higher temperatures were also found. No evidence of change in community care use was found. Evidence for the impact of heat on other mental health outcomes was mixed. High ambient temperatures have a range of mental health effects. The strongest evidence was found for increased suicide risk. Limited evidence was found for an increase in heat-related morbidity and mortality among people with known mental health problems. Knowledge gaps exist on the impact of high temperatures on many common mental health disorders. Mental health impacts should be incorporated into plans for the public health response to high temperatures, and as evidence evolves, psychological morbidity and mortality temperature thresholds should be incorporated into hot weather-warning systems.