Emergency maintenance request: what counts as an emergency, what is only urgent, and how to triage the 2am call

An emergency maintenance request is a condition that is actively damaging the property or endangering someone in it: uncontrolled water, a gas odor, no heat in freezing weather, no electricity, sewage backing up into the unit, fire or smoke, or a broken exterior door or lock that leaves a unit unsecured. Those get a vendor moving tonight. No hot water, a dead refrigerator, a sole toilet out of service, or air conditioning failing during a heat advisory are usually urgent rather than emergency: habitability is affected, but nothing is getting worse by the hour. Everything else is routine and belongs on the normal work order queue. The distinction matters because the repair clock in most states starts at notice, and notice is the phone call.

Last updated August 2026. Legal timelines below cite the statute or regulation directly. This is operational guidance for property managers and landlords, not legal advice for your jurisdiction.

What is considered an emergency maintenance request?

The working test most experienced operators use has two parts. Is the condition getting worse while nobody is doing anything about it, and would a reasonable person call this dangerous? A pipe that has already burst fails both halves of a wait-until-Monday argument, because every hour of running water compounds the damage and the mold exposure behind it. A dripping faucet fails neither.

Here is the three tier split written the way it should be handed to whoever answers your phone at 2am.

Two judgment calls sit inside that table and are worth deciding in advance rather than at midnight. The first is temperature. A no heat call in October in Georgia and a no heat call in January in Minnesota are the same sentence and completely different emergencies, which is why the most common written rule is a threshold: heat failure is an emergency when the forecast low is below a stated temperature, often somewhere around 40 to 50 degrees Fahrenheit, and urgent otherwise. Pick your number and write it down. The same logic applies to air conditioning in the other direction, and several states now treat extreme heat as a habitability issue in its own right.

The second is who is in the unit. A no heat call in a unit with an infant, an elderly tenant, or someone with a disclosed medical condition escalates regardless of the thermostat reading. That is not a legal technicality, it is basic risk management, and it should be an explicit override in your script rather than something you hope the person on the phone thinks of.

How long does a landlord have to fix an emergency repair?

Most states do not publish an hour count. They apply a reasonable time standard measured from the moment the landlord receives notice, and reasonable is read against how serious the condition is, which is a polite way of saying a court decides afterward. A handful of jurisdictions are concrete, and they are the useful reference points.

Massachusetts is the clearest example on the fast end. Under the State Sanitary Code at 105 CMR 410.630(A), a landlord must make a good faith effort to correct serious violations, the category that covers no heat, no hot water, and failed plumbing, within 24 hours of notice. Section 410.351 of the same code defines the hot water violation itself, treating a loss of adequate hot water lasting 24 hours or longer as a code condition.

California takes the opposite drafting approach. Civil Code section 1942(a) lets a tenant repair a condition that renders the premises untenantable and deduct the cost from rent if the landlord neglects the repair for a reasonable time after notice, capped at one month rent and available no more than twice in any twelve month period. Section 1942(b) then supplies a presumption: a tenant who acts after the 30th day following notice is presumed to have waited a reasonable time. That presumption is rebuttable in both directions, and the statute says explicitly that it does not stop a tenant from repairing and deducting sooner when the circumstances require shorter notice. A burst pipe is exactly the circumstance that requires shorter notice.

The practical lesson from both is the same and it has nothing to do with which state you are in. The clock starts at notice, so the single most valuable artifact you can produce is a timestamped record of when the tenant called, what they reported, and when you dispatched. A message pad does not produce that. A logged call does. If a habitability dispute ever surfaces, the difference between "we responded promptly" and a call log showing a 11:52pm report and a 12:19am vendor acknowledgement is the difference between an argument and a fact.

What temperature is considered an emergency maintenance request?

There is no national number, which is why this question keeps getting searched. What exists is a patchwork of local heat ordinances that set a minimum indoor temperature landlords must be able to maintain during a defined heating season, commonly in the 68 degree range during daytime hours, with a lower nighttime figure. Your city or state ordinance is the number that governs you, and it is worth looking up once and pinning to the wall.

For triage purposes, though, the ordinance is not the trigger. The trigger is the forecast. A heating failure becomes an emergency when the outdoor temperature over the next twelve hours will drive the unit below a habitable and safe range, and when pipes are at risk of freezing, which is its own property damage event separate from tenant comfort. Write the rule as a temperature threshold plus the vulnerable occupant override, give it to your answering service, and the 2am decision stops being a judgment call.

Is mold an emergency maintenance request?

Almost never as an emergency, and almost always as something you should not ignore either. Visible mold growth is not usually getting dramatically worse in the eight hours between a midnight call and a morning callback, so it does not meet the tonight test. What does meet it is the water that caused the mold. An active leak behind a wall is an emergency because it is compounding; the mold it produced last month is an urgent inspection. The right handling of a mold call at 1am is to establish whether water is currently running, treat that as the emergency if so, and book the inspection if not.

Handle the tenant conversation carefully in either case. Whoever answers your phone should record what was reported and schedule the inspection, and should not opine on whether the mold is dangerous, what species it is, or whether the unit is habitable. Those are determinations with legal consequences, and an offhand reassurance at 1am becomes an exhibit later.

What the call has to capture, or the dispatch fails anyway

Triage decides whether to send someone. Capture decides whether sending them accomplishes anything. Plenty of correctly classified emergencies still turn into a wasted trip charge because the person taking the call got the classification right and the details wrong.

• Property and unit number. Not the tenant name alone. Portfolios have repeat names and buildings have similar addresses.
• The specific failure, in enough detail to bring parts. "Water leak" sends a plumber with a guess. "Water coming from under the kitchen sink, cabinet is wet, shutoff valve is stuck" sends one with the right valve.
• Whether water, gas, electricity, or heat is involved. This is the field that decides the tier, so it should be asked directly rather than inferred.
• Occupancy right now, and access. Will the tenant be there? Is there a lockbox? Does the tech have permission to enter if nobody answers?
• Pets. A large dog behind the door is the most common reason a dispatched tech leaves without doing the work.
• A callback number, read back and confirmed. Confirmed matters. Transposed digits at 2am are extremely common.
• Timestamp. When the call came in and when the vendor acknowledged, because that pair is your record of response.

For commercial and mixed portfolios there is a preliminary question that residential managers rarely face: whether the repair is yours at all. Triple net and modified gross leases push varying slices of HVAC, roof, plumbing, and structural responsibility onto the tenant, and the answer sits in a clause somewhere in a fifty page document nobody wants to open at midnight. Getting a team to pull the repair and maintenance obligations out of every lease into a single reference sheet before the season turns is worth more than any script change, because the on-call person cannot triage a question they have no way to answer.

Write the rules down before you delegate them

The reason all of this matters operationally is that at some point the phone stops being yours. It goes to an on-call rotation, a junior coordinator, or an outside service, and every one of those handoffs transmits your judgment only as well as you have written it down. An answering service that has your three tier list verbatim, your temperature threshold, your vulnerable occupant override, your per-property on-call contacts, and your spend authorization limit will make the same decision you would. One that has none of those will guess, and it will guess conservatively at 2am, which is how a running toilet costs $350 in emergency call-out fees.

Five things belong in that document, and they take an afternoon to produce:

1. The three tier list, with your own examples, not a generic one.
2. The temperature threshold for heating and cooling failures, plus the vulnerable occupant override.
3. The on-call contact per property, per night, and the channel they actually watch, with a fallback contact and an acknowledgement window.
4. The spend threshold above which the service must reach you rather than authorize work.
5. The explicit do-not-say list: no repair timeline promises, no habitability opinions, no discussion of rent or damage responsibility.

That document is also the thing to hand any vendor you are evaluating, and their reaction to it tells you most of what you need to know. A service that can configure all five is running a rules engine. One that says it will note your preferences is taking messages. The gap between those two is the entire value of the category, and it is covered in more detail in our breakdown of what a property management answering service should do with a tenant call, including what the published US rate cards actually cost.

How the pricing model interacts with all of this

One last point that catches operators off guard. Emergency maintenance calls are long, because the capture list above takes time and the service often has to reach the on-call tech, hold, and call the tenant back. Volume is also weather driven rather than smooth: a hard freeze produces a night with more calls than the previous three weeks. Metered pricing bills that night at full rate and then at the overage rate once your included block is gone.

So when you compare vendors, divide the plan price by the included minutes and compare that number to the same vendor published overage rate, then ask what the billing increment is and whether hold time counts. PATLive publishes that it bills the first minute in full and then in six second increments, which is more disclosure than most of its competitors offer and is exactly the detail that decides your invoice. We work through the arithmetic across both billing structures in per minute versus per call answering service pricing, and the broader market rate card sits on our answering service pricing page. If the calls you are worried about are leasing inquiries rather than maintenance, the same structural questions come up in real estate answering service coverage, where the failure mode is a lost showing rather than a flooded unit.