In this quantitative study, the author examined the relationship between duration of hospitalization and frequency of assaultive behavior in 42 older long-term patients with dementia in a Canadian psychiatric hospital. The study instrument used for data collection was existing incident reporting forms routinely completed in Canadian regional psychiatric hospitals. A secondary analysis was conducted using data previously collected on a regular basis by the psychiatric hospital serving as the study site. A significant negative correlation was found between the number of assaults committed and the number of months spent in the hospital, with significantly fewer assaults occurring in the second year of hospitalization compared with the first year. Male patients were observed to be significantly more assaultive than female patients. Findings suggest that the maximum benefit for patients hospitalized for assaultive behavior is obtained during the first 2 years of inpatient treatment and that patients within this population who are no longer assaultive may be more appropriately cared for in nursing homes. Based on these findings, resources should be allocated to assist with the transition of formerly assaultive patients with dementia from a psychiatric hospital to a nursing home. This scenario forecasts the development of a challenging new role for nurses.