Jun 9, 2026 · ShiftFlow Editorial Team · Compliance · 9 min read

How On-Call and Standby Pay Works for Emergency Service Calls

A tech carrying the after-hours phone is not automatically on the clock — but tighten the restrictions enough and that standby time becomes payable. Here is where federal law draws the line, and why the dispatch timestamp matters.

Carrying the Phone Isn’t Automatically Working

Every service business with emergency coverage runs into the same question. A tech has the after-hours phone this week. Burst pipe at 11 PM, no-heat call at 2 AM — that part is obviously paid. But the other twenty-some hours a day, when the tech is just carrying the phone, waiting, living their life? Do you owe wages for that?

It’s a real money question. Pay every standby hour and an on-call week costs you a fortune. Pay none of them and, if your on-call rules are strict enough, you may be underpaying in a way that turns into back pay. Federal law has a framework for this. Like most wage questions, it’s less about the label “on call” and more about how much the arrangement actually restricts the tech.

This is general information, not legal advice. On-call pay is genuinely fact-specific — the outcome turns on the total picture of a given setup, and several states are stricter than the federal floor.

The Line — Engaged to Wait vs Waiting to Be Engaged

The whole area runs on one distinction. Waiting time is paid when the worker is “engaged to wait” — the waiting is part of the job and the time belongs to the employer. It’s unpaid when the worker is merely “waiting to be engaged” — completely relieved of duty long enough to use the time for their own purposes. That framework comes from 29 CFR 785.14 through 785.16.

A tech standing around at a customer site waiting for a unit to depressurize is engaged to wait. Paid. A tech at home with a phone, free to watch the game and unlikely to be called, is closer to waiting to be engaged. Unpaid. The on-call rule is just this distinction applied to standby.

The On-Call Rule

The specific regulation is 29 CFR 785.17, and it sets up a clean contrast:

A tech required to stay on the employer’s premises while on call is working — that time is paid. Bright-line.
A tech on call at home, or who just has to leave word where they can be reached, is generally not working while on call — that time is generally unpaid.

For most service businesses, the second case is the norm: the tech takes the phone home, lives normally, and gets paid when a call actually comes in. The DOL’s Fact Sheet #22 states this directly, then adds the crucial caveat that “additional constraints on the employee’s freedom could require this time to be compensated.” That caveat is where the disputes live.

When Restrictions Flip Standby to Paid

At-home on-call becomes payable when the restrictions you impose are severe enough that, taken together, the tech can’t really use the time for themselves. There’s no single trigger. Courts weigh the totality of the circumstances, and no one factor decides it. Here are the factors that come up repeatedly.

Required response time. How fast must the tech be on site or working? This is often the most telling factor. A window so short it effectively chains the tech to the truck points toward paid; a comfortable one-to-two-hour window points toward unpaid. (These are guideposts from case law, not statutory thresholds — do not treat any specific number as a magic line.)
Geographic restrictions. Hard limits on how far the tech can be from the service area cut against personal use.
Call frequency. If calls come so often that the standby period is constantly interrupted, the time starts to look like the employer’s, not the tech’s.
Ability to trade shifts. If a tech can easily hand off the on-call duty, that weighs toward unpaid.
Whether a phone frees them up. Being reachable by phone — able to roam, shop, eat out — weighs toward unpaid, because the tech can still live their life.

Courts have gone both ways depending on the mix. A 20-minute required response with several calls a shift has been found compensable; a similar response window where the tech could still shop, dine, and pursue leisure has been found not compensable. The lesson for a contractor isn’t a formula. It’s that the tighter you make the leash, the more likely the standby hours are payable. If you need fast response, you may simply need to pay for the standby — a cost to price into emergency service rather than a rule to engineer around. And subjective stress alone doesn’t make it paid. The restriction has to be objectively severe.

Once the Call Comes In, the Clock Is Running

Here the rule is clean, not gray. When an on-call tech is actually dispatched and starts working, that time is compensable hours worked — and it counts toward the 40-hour workweek. If the 2 AM call-out pushes the tech over 40 hours for the week, those overtime hours are paid at one and a half times the regular rate, under the FLSA’s overtime provisions.

This is exactly where accurate records matter most, because the pay hinges on two timestamps: when the tech was dispatched and started the response, and when they finished. Whether the drive to the emergency is fully paid overlaps with ordinary travel-time rules and can itself be fact-specific, but the on-site work time is squarely payable. A 90-minute call-out at 2 AM that nobody logs precisely becomes a guess at payroll. And guesses about overtime hours, at time-and-a-half, are expensive in both directions.

The Standby Stipend Trap

Many service businesses pay a flat stipend for carrying the phone — say a set amount per on-call shift — even when the bare standby hours aren’t legally required to be paid. It’s good for morale and recruiting. But it has an overtime consequence people miss.

A prearranged, regularly paid standby stipend is generally not one of the statutory exclusions from the “regular rate” of pay. That means it typically has to be folded into the regular-rate calculation when you compute overtime: you add the stipend to the week’s other pay, divide by total hours worked to get the regular rate, and pay the overtime premium on that slightly higher rate. So a $50-per-shift on-call stipend doesn’t just cost you $50. In any week the tech works overtime, it nudges the overtime rate up too. (Premiums for genuinely unscheduled emergency call-backs can be treated differently, but a scheduled, prearranged standby stipend doesn’t get that treatment.) This is worth confirming with whoever runs your payroll, because a wrong regular rate quietly underpays every overtime hour.

The California Caution

As with most wage rules, the federal standard is a floor, and California is stricter. California uses a “control” test rather than the federal duty test, and it can require pay for on-call time the FLSA would treat as unpaid — the California Supreme Court has expressly declined to import the more lenient federal standard for on-call and on-site standby. California also has separate reporting-time pay rules. Other states have their own variations. If you operate in California or any state with its own wage rules, don’t assume the federal answer governs.

Why It Comes Down to Your Records

Step back and notice what every part of this depends on: knowing exactly when standby started and ended, when a tech was dispatched, when the call-out finished, and how it all adds up against the 40-hour line. The legal analysis is yours (and your attorney’s) to make. But you can’t apply it to hours you didn’t capture.

A time clock built for the field is the record-keeper here. A tech can log the moment they’re dispatched, track the call-out as its own block tied to the customer, and have those after-hours hours roll into the week’s overtime total automatically — so the 2 AM call is paid correctly and the overtime math is right. If you pay standby stipends, keeping the call-out hours cleanly separated from the stipend is what lets payroll compute the regular rate properly. ShiftFlow, a time clock built for service trades, captures and timestamps those events and exports them to payroll as PDF or CSV. It doesn’t decide which standby hours are legally compensable or run the regular-rate calculation for you. Those are your calls. What it removes is the part where you’re reconstructing a month of midnight call-outs from memory and hoping the overtime came out right.

Frequently Asked Questions

Do you have to pay a tech just for being on call?

Under federal law, generally not, if the tech is on call at home or simply has to be reachable and is otherwise free to use the time for personal purposes. That’s treated as waiting to be engaged, which is unpaid. It flips to paid when the restrictions are severe enough that the tech can’t effectively use the time for themselves: a very short required response window, tight geographic limits, or calls so frequent the time isn’t really their own.

When does on-call time become paid?

When the restrictions placed on the tech are heavy enough that, taken together, they prevent the worker from using the time effectively for personal activities. Courts weigh factors like required response time, geographic limits, call frequency, the ability to trade shifts, and whether a phone lets the worker move around freely. No single factor decides it; it’s the total picture. On-call time spent on the employer’s premises is paid.

Is the time a tech spends on an actual emergency call paid?

Yes. Once an on-call tech is dispatched and starts working, that time is compensable hours worked, and it counts toward the 40-hour workweek. If the call-out pushes the tech past 40 hours, the overtime hours are paid at one and a half times the regular rate. Capturing the exact dispatch and finish times is what makes that pay accurate.

Does a flat on-call stipend count toward overtime?

Usually yes. A prearranged, regularly paid standby stipend is generally not one of the statutory exclusions from the regular rate, so it typically has to be folded into the regular-rate calculation for overtime. That means the stipend can raise the overtime rate for any week the tech works more than 40 hours. The treatment of true unscheduled call-back premiums can differ, so confirm the specifics.