The present study evaluates data from 116 forensic inpatients who underwent violent risk assessments, which included the Historical, Clinical, Risk‐20 (HCR‐20), from 2006 to 2013 as part of an opportunity to be conditionally discharged from state forensic facilities. Of the 116 inpatients, 58 were never released, 39 were released and returned to a hospital, and 19 were released and never returned. Results from analyses of variance and multinomial logistic regression found the risk management (R) scale of the HCR‐20 successfully predicted group membership in that higher scores were associated with a greater likelihood of not being released from a forensic facility or returning to a forensic facility after release. The results of this study indicate that clinicians should consider community‐based risk variables when evaluating forensic patients for potential return to the community. This research demonstrates that clinicians failing to fully consider dynamic risk factors associated with community integration jeopardize the quality and thoroughness of their violence risk assessment with regards to readiness for release. Copyright © 2016 John Wiley & Sons, Ltd.