Critical illness and its treatment often result in long-term neuropsychiatric morbidities. Consequently, there is a need to focus on means to prevent or ameliorate these morbidities. Animal models provide important data regarding the neurobiological effects of physical activity, including angiogenesis, neurogenesis, and release of neurotrophic factors that enhance plasticity. Studies in noncritically ill patients demonstrate that exercise is associated with increased cerebral blood flow, neurogenesis, and brain volume, which are associated with improved cognition. Clinically, research in both healthy and diseased human subjects suggests that exercise improves neuropsychiatric outcomes. In the critical care setting, early physical rehabilitation and mobilization are safe and feasible, with demonstrated improvements in physical functional outcomes. Such activity may also reduce the duration of delirium in the intensive care unit (ICU) and improve neuropsychiatric outcomes, although data are limited. Barriers exist regarding implementing ICU rehabilitation in routine care, including use of sedatives and lack of awareness of post-ICU cognitive impairments. Further research is necessary to determine whether prior animal and human research, in conjunction with preliminary results from existing ICU studies, can translate into improvements for neuropsychiatric outcomes in critically ill patients. Studies are needed to evaluate biological mechanisms, risk factors, the role of pre-ICU functional level, and the timing, duration, and type of physical activity for optimal patient outcomes.