A substantial body of evidence supports the role of exercise interventions for people with a mental illness. However, much of this literature is conducted using outpatient and community-based populations. We undertook a systematic review examining the effect of exercise interventions on the health of people hospitalized with depression, schizophrenia, bipolar disorder, or anxiety disorders. Eight studies met our inclusion criteria. Several studies show positive health outcomes from short-term and long-term interventions for people hospitalized due to depression. Although positive, the evidence for inpatients with schizophrenia, bipolar disorder, or anxiety disorders is substantially less. There is an urgent need to address the paucity of literature in this area, in particular the optimal dose and delivery of exercise for people hospitalized as a result of mental illness. Standardization of reporting exercise programme variables, the assessment of mental illness, and the reporting of adverse events must accompany future studies.