HMIS exit destination categories

The same list of living situations does three jobs in HMIS. It answers where somebody was before they came to you, where they are during an outreach contact, and where they are going when they leave. That last use is the destination, and it carries more weight than the other two because it is what your outcomes are made of.

The rule that makes the list readable

Every option is numbered, and the first digit tells you which category it belongs to. Once you know that, you never have to memorise the list.

100 to 199 Homeless situations Still homeless on exit. Never a positive outcome.

200 to 299 Institutional situations Hospital, jail, treatment, foster care, nursing home. Neither homeless nor housed.

300 to 399 Temporary situations Somewhere to sleep that is not somewhere to stay. Not a permanent destination.

400 to 499 Permanent situations The ones that count. This is what a positive housing outcome is made of.

1 to 99 Everything else Other, deceased, no exit interview completed, worker unable to determine, client doesn't know, client prefers not to answer, data not collected. Read the section on these before using any of them.

So a destination in the 400s is a housing outcome and a destination in the 100s is not, and you can tell at a glance without looking anything up. Place not meant for habitation is 116. Emergency shelter is 101. Safe haven is 118. All homeless, all hundreds.

What is in each category

Homeless 100s	Place not meant for habitation, such as a vehicle, an abandoned building, a station or outside · Emergency shelter, including a hotel or motel paid for with an emergency shelter voucher, or a Host Home shelter · Safe Haven
Institutional 200s	Foster care home or foster care group home · Hospital or other residential non-psychiatric medical facility · Jail, prison or juvenile detention facility · Long-term care facility or nursing home · Psychiatric hospital or other psychiatric facility · Substance abuse treatment facility or detox center
Temporary 300s	Transitional housing for homeless persons, including homeless youth · Residential project or halfway house with no homeless criteria · Hotel or motel paid for without an emergency shelter voucher · Host Home, non-crisis · Staying or living with family, temporary tenure · Staying or living with friends, temporary tenure · Moved from one HOPWA funded project to HOPWA transitional housing
Permanent 400s	Staying or living with family, permanent tenure · Staying or living with friends, permanent tenure · Moved from one HOPWA funded project to HOPWA permanent housing · Rental by client, no ongoing housing subsidy · Rental by client, with ongoing housing subsidy · Owned by client, no ongoing housing subsidy · Owned by client, with ongoing housing subsidy

Individual code numbers live in your HMIS dropdown and in HUD's Appendix A, which is the authority and is kept current. We point at it rather than copying it, because a copied code table goes stale quietly and this is not a page you want to be quietly wrong.

The row where most agencies get it wrong

Look at the temporary and permanent categories above. Staying with family appears in both. Staying with friends appears in both. Same couch, same address, same person. One version sits in the 300s and is not a housing outcome. The other sits in the 400s and is.

The only thing separating them is tenure, meaning whether the person can stay. Not whether they want to stay, not whether it is a good situation, and not whether you approve of it. Whether the arrangement is understood by everyone involved to be ongoing.

Why this row matters more than the rest of the list

Doubling up with family is one of the most common places people go when they leave a program. It is also the row where a rushed exit interview does the most damage, because the difference between the two options is a question somebody has to actually ask.

An agency that defaults to temporary understates its outcomes. An agency that defaults to permanent overstates them and will see it come back through returns to homelessness two years later. Neither default is a decision anyone made on purpose.

HUD's own guidance is that the destination should reflect where the person is expected to stay, that it should come directly from the client rather than from a secondhand source, and that it should not be an assumption carried forward from what was recorded at enrollment.

A question that works: "Is this somewhere you can stay as long as you need to, or is it more of a for-now thing?" It is answerable, it does not make anyone defend their situation, and it maps cleanly onto the two options.

The 1 to 99 codes, and why to avoid most of them

These are the escape hatches, and every one of them costs you something.

OtherHUD says to use it as a last resort, and it does not count as a positive outcome. If your reports show a meaningful share of exits to Other, that is usually a training gap rather than a population that genuinely fits nowhere on a list this long.

No exit interview completedHonest and sometimes unavoidable. People leave without saying goodbye. It still lands as a non-permanent exit in every calculation, so it is worth knowing how often it happens and whether the reason is genuinely that people vanished.

Worker unable to determineDifferent from the one above. Here you had contact but could not establish the destination. Real, and it should be rare.

Client doesn't know, client prefers not to answerBoth are legitimate answers a person is entitled to give, and both are better than a guess. Prefers not to answer in particular is a client exercising a choice, and recording it accurately is the respectful option as well as the correct one.

Data not collectedMeans nobody asked. It is the only one on this list that is purely a process failure, and it is the one to drive toward zero.

Where the destination shows up later

One field, recorded once, in a hurry, at the end of a case. It reappears in at least four places:

Your APR, in the exit destination questions, and in the versions of them broken out by race and ethnicity.
System Performance Measure 7, the rate of successful placement, which is one of the metrics the CoC Application asks about directly.
System Performance Measure 2, returns to homelessness, where a permanent destination starts a two year clock on whether it held.
Your local competition score, since at least a quarter of the available points must be based on system performance measures.

A blank destination does not sit out any of those calculations. It counts as a placement that was not permanent.

About COMPASS

We make case management software for organizations doing this work. It captures the destination as part of the exit a worker is already recording, and names the required fields missing from a record before an export is generated.

That is the only pitch on this page. The rest is true whatever system you run.

Got a destination that does not fit any of these?

Send it over. Those edge cases are worth collecting and we will add the good ones to this page.

hello@illumipath.io

Sources

Labels here follow HUD's Appendix A. For the individual code numbers use Appendix A directly or your own HMIS, both of which stay current in a way a copied table on someone else's website does not.