Objective Low frequency of going outdoors (e.g. being homebound) is associated with depressive mood; however, the underlying neural mechanism of this association is unclear. We therefore investigated the neural substrate involved in the relationship between frequency of going outdoors and depressive mood using positron emission tomography (PET), focusing on the frontal lobe and the limbic system. Methods One hundred fifty‐eight community‐dwelling older adults aged 65–85 years underwent PET with 18F‐fluorodeoxyglucose to evaluate regional cerebral metabolic rates of glucose normalized in reference to cerebellar glucose metabolic value (normalized‐rCMRglc) in six regions of interest. We also assessed depressive mood, frequency of going outdoors, and potential covariates. Depressive mood was assessed using the Geriatric Depression Scale (GDS). Results The proportion of participants who reported low frequency of going outdoors (LG, every 2–3 days or less) was 36.1%. The LG group showed significantly higher GDS scores than those who reported high (once a day or more) frequency of going outdoors. A multiple linear regression analysis adjusted for potential covariates showed higher GDS scores were associated with lower normalized‐rCMRglc in the ventrolateral prefrontal and orbitofrontal cortices. Adjusting for frequency of going outdoors, the association between GDS score and normalized‐rCMRglc in the orbitofrontal cortex was attenuated. Conclusions Our results suggest that the orbitofrontal cortex may mediate the relationship between low frequency of going outdoors and depressive mood among community‐dwelling older adults. These findings may help disentangle the role of going outdoors in regulating brain function to improve and/or maintain mental health among community‐dwelling older adults. Copyright © 2016 John Wiley & Sons, Ltd.