Introduction. Unsheltered homelessness is an important phenomenon yet difficult to study due to lack of data. The Veterans Health Administration administers a universal homelessness screener, which identifies housing status for Veterans screening positive for homelessness. Methods. This study compared unsheltered and sheltered Veterans, assessed differences in rates of ongoing homelessness, and estimated a mixed-effect logistic regression model to examine the relationship between housing status and ongoing homelessness. Results. Eleven percent of Veterans who screened positive for homelessness were unsheltered; 40% of those who rescreened were homeless six months later, compared with less than 20% of sheltered Veterans. Unsheltered Veterans were 2.7 times as likely to experience ongoing homelessness. Discussion. Unsheltered Veterans differ from their sheltered counterparts— they are older, more likely to be male, less likely to have income—and may be good candidates for an intensive housing intervention. Future research will assess clinical characteristics and services utilization among this population.