On-Call Pay Rules: When Standby Time Is Compensable

On-call time is paid when the employee is not really free to use the time as their own.

That is the practical rule for employers. A worker who only keeps a phone nearby may be unpaid for standby time. A worker who must stay on site, answer within minutes, stay sober and nearby, or sleep at the job may be working even when no call comes in.

Federal law decides the middle cases under the totality of the circumstances. The doctrine comes from Skidmore v. Swift & Co., 323 U.S. 134 (1944): courts look at how much control the employer has over the employee's time.

California is stricter, especially for 24-hour duty:

• Federal: 29 CFR §785.22 permits a sleep-time deduction for shifts of 24+ hours — up to 8 hours per shift, if there's a written agreement, adequate sleeping facilities, and the employee usually enjoys an uninterrupted night's sleep.
• California: Mendiola v. CPS Security Solutions, Inc., 60 Cal.4th 833 (2015) held that on-premises 24-hour security guards are owed ALL 24 hours — the federal §785.22 sleep-time deduction does NOT carry over to California state-law claims.

The biggest on-call exposure category is healthcare: residents, ER physicians, hospital staff, and caregivers. The fastest-growing is IT and DevOps: PagerDuty rotations, system alerts, and 5- to 15-minute response expectations. The federal floor is 29 CFR §785.17:

"An employee who is required to remain on call on the employer's premises or so close thereto that he cannot use the time effectively for his own purposes is working while 'on call.' An employee who is not required to remain on the employer's premises but is merely required to leave word at his home or with company officials where he may be reached is not working while on call."

The bright-line rule sits at the two poles. Real cases sit in the middle: a 30-minute response window with an alcohol restriction, a PagerDuty acknowledge-in-15-minutes expectation, or a hospital resident with a pager and a 20-mile geographic limit. Those are fact questions.

This guide covers when standby time is paid, when sleep time can be deducted, why California is different, how stipends affect overtime, and what records an employer needs when an on-call dispute turns into a wage claim.

Quick reference

• Federal floor (29 CFR §785.17): on-call ON the employer's premises = compensable. On-call merely "leave word where you can be reached" = NOT compensable. Everything in between is Skidmore totality.
• The factors: on-premises vs off-premises restriction, required response time, geographic limits, frequency of calls, ability to engage in personal activities, ability to trade shifts. No single factor is dispositive.
• Engaged to wait (waiting is part of the job, employer benefit) → compensable. Waiting to be engaged (free to use the time) → NOT compensable. §785.16 + Skidmore.
• 24-hour duty federal rule (§785.22): up to 8 hours sleep deduction, if there's a written agreement + adequate facilities + usually-uninterrupted sleep. Interruptions = hours worked. Less than 5 consecutive sleep hours = entire scheduled sleep period is compensable.
• California (Mendiola): the §785.22 sleep-time deduction does NOT apply to California state-law claims. On-premises 24-hour duty = all 24 hours compensable.
• Less-than-24-hour duty (§785.21): entire on-duty period is hours worked, sleep facilities or not.
• Healthcare 8-and-80 rule (FLSA §7(j)): hospitals can compute overtime over a 14-day period (8 hours/day + 80 hours/14 days) instead of standard 40-hour workweek — changes the regular-rate math when on-call gets pulled in.
• Public safety (FLSA §7(k)): police and fire on-call rules sit under a separate exemption with 7- to 28-day work periods.

The 5 Most Expensive On-Call Mistakes

Before the rule mechanics, the patterns that drive litigation. Each is a routine class-action target.

1. Treating an on-call stipend as the entire compensation. A flat $200/week on-call stipend doesn't cover compensable on-call hours under Skidmore. If the totality test makes the standby time hours-worked, the stipend is just supplemental — the employee is still owed the hourly rate × the compensable hours, and the stipend is included in the regular rate for overtime under 29 CFR §778.208–§778.209. Many tech and healthcare employers pay a stipend and stop there, then discover the underlying compensable-hours liability years later.

2. The Mendiola trap: 24-hour security duty in California. Mendiola v. CPS Security Solutions, 60 Cal.4th 833 (2015) made on-premises 24-hour security guard on-call time fully compensable in California — and rejected the federal §785.22 sleep-time deduction for state-law claims. Multi-state security operations with uniform "we pay 8 of the 24 hours" policies have material California exposure even when the policy is defensible federally.

   Cited cases

   • Mendiola v. CPS Security Solutions, Inc., 60 Cal.4th 833 (2015)
     California Supreme Court — On-premises 24-hour security-guard on-call fully compensable under CA "subject to control" test; federal §785.22 sleep-time deduction does NOT apply to state-law claims
   • Troester v. Starbucks Corp., 5 Cal.5th 829 (2018)
     California Supreme Court — California rejects federal de minimis doctrine; brief working-while-on-call acknowledgments are compensable

3. Failing the §785.22 sleep-time-deduction conditions for 24+ hour duty. The deduction requires (a) written agreement, (b) adequate sleeping facilities, (c) employee usually enjoys an uninterrupted night's sleep. Most employers get the agreement (a handbook clause). Many get the facilities (a bed in a quiet room). Almost none track the actual interruption frequency — and when call logs surface in litigation showing nightly interruptions, the deduction collapses and the full 8 hours retroactively become compensable.

4. PagerDuty / Opsgenie / ServiceNow on-call rotations without explicit Skidmore analysis. IT and DevOps on-call patterns have 5–15 minute response SLAs and tracked acknowledgment times. The shorter the SLA and the more frequent the calls, the more compensable the standby time under Skidmore. Most tech employers pay a small on-call stipend + hourly rate on actively-engaged minutes; the open exposure is the standby time the employee can't really use because they're waiting for the next page.

5. Healthcare on-call layered with the §7(j) 8-and-80 rule. Hospitals using the §7(j) hospital exemption compute overtime over 14-day periods (8/day + 80/14-days). Compensable on-call hours pulled into that period change the overtime math in ways most payroll systems handle as a flat addition. Hospital wage-and-hour class actions in 2024–2026 have routinely surfaced sleep-time + §7(j) interactions that aggregate into seven-figure exposures.

   Cited cases

   • Skidmore v. Swift & Co., 323 U.S. 134 (1944)
     US Supreme Court — Totality-of-circumstances test for on-call and waiting time; question of fact decided case-by-case
   • Armour & Co. v. Wantock, 323 U.S. 126 (1944)
     US Supreme Court — Companion case applying the same totality test to private firefighters at Armour soap factory

Federal Baseline: 29 CFR Part 785 and the Skidmore Test

The broader pre/post-shift framing — security checks, donning/doffing, after-hours messaging — sits in our off-the-clock work laws by state guide. The compensable-travel categories under §§785.33–.41 live in our travel time pay rules guide. This article focuses on the §785.16–.22 on-call and waiting-time rules.

29 CFR §785.17 — On-call time (verbatim)

"An employee who is required to remain on call on the employer's premises or so close thereto that he cannot use the time effectively for his own purposes is working while 'on call.' An employee who is not required to remain on the employer's premises but is merely required to leave word at his home or with company officials where he may be reached is not working while on call."

Two poles. The bright-line rule covers the easy cases:

• Required to remain on the premises → working (compensable).
• Merely required to leave word → not working (not compensable).

Everything between those poles — restricted off-premises on-call, mandatory response windows, geographic limits — gets decided by the Skidmore totality test under §785.16.

29 CFR §785.16 — Off duty (waiting time, engaged-to-wait distinction)

"Periods during which an employee is completely relieved from duty and which are long enough to enable him to use the time effectively for his own purposes are not hours worked."

The companion rule. The doctrinal distinction:

• "Engaged to wait" — waiting is part of the job; the primary benefit runs to the employer; the employee is on duty even when nothing is actively happening. Compensable.
• "Waiting to be engaged" — the employee is free to use the time as their own, with only minor responsiveness obligations. Not compensable.

§785.16 itself is short. The substantive doctrine comes from the 1944 Supreme Court companion cases.

Skidmore v. Swift & Co., 323 U.S. 134 (1944)

Facts. Skidmore and six co-workers were firefighters at Swift's meatpacking plant. They worked daytime shifts five days a week. Three-and-a-half to four nights a week, they stayed overnight in the fire hall on the plant's premises to respond to alarms. They had beds, could eat and read and play cards, and were paid only for time spent actually responding to alarms. They sued for compensation for all the overnight hours.

The Supreme Court held: there's no legal formula. Whether on-call time is "working time" under the FLSA is a question of fact for the trial court. The court must scrutinize the employment agreement, appraise its practical construction, consider the nature of the service, and weigh "all of the surrounding circumstances."

The frame: totality of the circumstances. The Court explicitly rejected a bright-line rule. Skidmore has controlled every on-call dispute since.

Armour & Co. v. Wantock, 323 U.S. 126 (1944) — the companion case

The same Supreme Court day issued Armour & Co. v. Wantock, applying the same totality test to private firefighters at Armour & Co.'s soap factory. The pairing matters: Skidmore + Armour together define the federal on-call framework. Modern federal appellate decisions on on-call routinely cite both.

The factors courts weigh (post-Skidmore federal case law)

Lower federal courts have distilled the Skidmore totality into a working list of factors. No single factor is dispositive, but the more boxes that check toward "employer controls the time," the more likely the on-call time is compensable:

• On-premises vs off-premises. On-premises strongly favors compensability; off-premises shifts the analysis to the other factors.
• Required response time. A 5-minute response window is far more restrictive than a 30-minute or one-hour window. Recent appellate decisions have found 15-minute response windows compensable in tech and healthcare contexts.
• Geographic limits. A "stay within 20 miles of the hospital" or "stay within the city limits" restriction limits what the employee can do during the standby period.
• Frequency and unpredictability of calls. Calls every hour ≠ calls once a week. The more frequent and unpredictable, the less the employee can meaningfully use the time.
• Personal-activity restrictions. Prohibitions on alcohol, on certain activities (driving, traveling, sleeping in noisy environments), or on having anyone in the home shift the analysis.
• Ability to trade or substitute shifts. When an employee can hand off the on-call to a colleague, the time is less restrictive than when they can't.
• Frequency of actual interruption (especially for 24-hour duty). Nightly call interruptions defeat the §785.22 sleep-time deduction even when all the formal conditions are met.

Federal circuits apply these factors slightly differently, but the framework is uniform. The 9th Circuit's Berry v. County of Sonoma, 30 F.3d 1174 (9th Cir. 1994), is the leading appellate application — and it's defendant-favorable (see below).

Berry v. County of Sonoma, 30 F.3d 1174 (9th Cir. 1994)

Facts. Sonoma County coroner deputies were on call 24 hours a day, seven days a week. They were required to answer a page or telephone call within 15 minutes and respond to death reports as soon as possible.

The Ninth Circuit held the on-call time was NOT compensable under the FLSA. Applying the Skidmore totality test, the court found that the 15-minute response requirement and other restrictions did not predominantly benefit the employer to the degree required for the standby time to be hours worked. Deputies could engage in most personal activities, sleep at home, eat with family, and trade calls with colleagues. The Ninth Circuit reversed the district court's contrary ruling.

Why it matters. Berry is the federal-floor benchmark for restricted off-premises on-call. A 15-minute response window, on these facts, was not enough to make the federal totality test flip toward compensability. Plaintiffs' counsel respond by raising state-law claims under broader-than-FLSA frameworks (CA, NY, WA) — the federal floor isn't the only game.

29 CFR §785.21 — Less than 24 hours on duty

"If an employee is required to be on duty for less than 24 hours, he is working even though he is permitted to sleep or engage in other personal activities when not busy."

A straight bright-line rule: under 24 hours, all on-duty time is hours worked. No sleep-time deduction is available. The telephone-operator example in the regulation makes the rule explicit: an operator on a 16-hour shift permitted to nap between calls is working the entire 16 hours.

29 CFR §785.22 — 24 hours or more on duty (sleep-time deduction)

For shifts of 24+ hours, the employer and employee may agree to exclude bona fide regularly scheduled sleeping periods of not more than 8 hours, provided:

1. The agreement is express (written or implied by practice).
2. Adequate sleeping facilities are furnished.
3. The employee can usually enjoy an uninterrupted night's sleep.

Interruption rules. Any actual interruption of the sleep period is hours worked. If interruptions prevent the employee from getting at least 5 hours of consecutive sleep, the entire scheduled sleep period becomes compensable.

Worked example. A hospital schedules a 24-hour resident shift with an 8-hour scheduled sleep period (11pm–7am), satisfying the written-agreement and adequate-facilities conditions. On Monday the resident is paged at 11:45pm (15 min), 2:30am (20 min), and 5:15am (45 min) — three interruptions totaling 80 minutes. The longest consecutive sleep window is 2 hours 30 minutes (from 12:00am to 2:30am). Because the resident did not get at least 5 hours of consecutive sleep, the entire 8-hour scheduled sleep period is compensable — not just the 80 minutes of actual interruption. At a $50/hour resident rate, that's $400 in retroactive wages owed for one night, on top of the active-engagement pay already covered. Compounded across a 28-day resident rotation with 4 such interrupted nights per week (the realistic call-load for many programs), the per-resident exposure crosses $6,400 — and the §216(b) liquidated-damages doubling makes that $12,800 if the violation is found willful.

Practical application. This is the load-bearing federal regulation for healthcare 24-hour duty, security-guard 24-hour shifts (the Mendiola pattern), EMS, fire, and other public safety. The four conditions are easy to claim and hard to maintain — most employers get the agreement and facilities right, then fail the "usually enjoys an uninterrupted night's sleep" condition when actual call logs surface in litigation.

California — Broader Than FLSA

California's IWC Wage Orders define "hours worked" as any time the employee is "subject to the control of an employer" — broader than the federal Skidmore totality test. The state Supreme Court has explicitly rejected the federal §785.22 sleep-time deduction for state-law claims, and the post-2018 Troester no-de-minimis rule compounds the exposure for small working-while-on-call moments.

Mendiola v. CPS Security Solutions, Inc., 60 Cal.4th 833 (2015)

Facts. CPS Security employed guards at California construction sites under a schedule of 16-hour shifts (8 hours on duty + 8 hours on-call) on weekdays and 24-hour shifts (16 hours on duty + 8 hours on-call) on weekends. Guards were required to live in employer-provided trailers at the work sites during their shifts and to stay within 30 minutes' response distance even during on-call hours. CPS counted only the active-duty hours as compensable; the on-call hours were unpaid.

The California Supreme Court held: ALL on-call hours are compensable under California IWC Wage Order 4. The court found CPS exercised "substantial control" over guards during on-call hours (mandatory on-site residency, 30-minute response distance, alarm-response obligation), and the federal §785.22 sleep-time deduction safe harbor does NOT apply to California state-law claims.

Practical effect. Any California employer with on-premises 24-hour-duty on-call patterns — security, residential care, some healthcare — owes the full on-call hours. The federal sleep-time deduction is unavailable for California claims regardless of facilities, written agreement, or actual interruption rates. Class-action settlements in the wake of Mendiola run from $1M to $10M+ for medium-sized security workforces.

Troester v. Starbucks — applied to on-call

Troester v. Starbucks Corp., 5 Cal.5th 829 (2018) rejected the federal de minimis doctrine for California state-law claims. Applied to on-call: any working-while-on-call activity that's compensable on the merits — answering a 30-second call, acknowledging a 15-second PagerDuty alert — is owed in California, no matter how small. The federal ~10-minute-per-day de minimis floor doesn't apply.

Things California employers consistently miss

• On-premises 24-hour duty without the §785.22 safe harbor. A national security firm with one California site running 24-hour rotations under a federal sleep-time policy will discover the Mendiola exposure first when a class action is certified. The fix is either to eliminate the on-call portion of the schedule or to pay all 24 hours in California regardless of duty status.
• PagerDuty / Slack acknowledgment time as compensable working-while-on-call. The federal de minimis defense (under §785.47) routinely excuses small per-event compensable time. California does not. Acknowledging a 20-second page is Troester-compensable in California.
• Stipend-only compensation for restricted on-call. A $200/week on-call stipend doesn't cover compensable hours in California any more than it does federally — and California's broader "subject to control" test makes more on-call time compensable than Skidmore would. The stipend must be included in the regular rate for overtime under §778.208–§778.209, and the underlying hourly rate is owed on top for compensable standby time.
• Mandatory call-trading restrictions. When an employee can't trade an on-call shift with a colleague — and call-trading is prohibited by policy — the employee's control over their own time is reduced. Skidmore federal courts weigh this as one factor among many; California Mendiola-style analysis treats it as direct evidence of "subject to control."

State-by-state — beyond the federal floor

California is the strictest, but several states have hours-worked definitions that go beyond the federal Skidmore totality framework. The table covers the states where state-law on-call exposure differs meaningfully from the federal baseline.

Industry-Specific Patterns

On-call exposure is concentrated in industries where the standby-and-respond model is structural. Each has its own characteristic litigation pattern.

Healthcare — residents, ER, hospital staff

The largest on-call exposure category in the country. Three sub-patterns:

• Hospital residents on home call with pager response. Restricted to 20–30 minutes of the hospital, alcohol prohibited, geographic limits. Federal courts have split on whether resident on-call is compensable under Skidmore; CA, NY, and WA state law typically broader.
• ER physicians with on-call coverage. Generally pager-based; most courts apply Skidmore to find off-premises on-call NOT compensable absent severe restriction.
• 24-hour inpatient staff (nurses, residents, hospital floor staff). §785.22 sleep-time deduction permissive federally; defeated when interruptions are frequent. In California, Mendiola governs and the deduction is unavailable.

The hospital §7(j) 8-and-80 overtime rule (29 USC §207(j) + 29 CFR §778.601) layers on top of the on-call analysis: when on-call hours are compensable, they enter the 14-day overtime calculation, changing the regular rate and the overtime owed. Most hospital payroll systems handle this poorly.

IT support and DevOps

The fastest-growing on-call exposure category. PagerDuty, Opsgenie, ServiceNow, and similar systems run rotations with 5–15 minute response SLAs. The platforms track exact acknowledgment times, response times, and resolution times — the same kind of exact-time records that drive Mt. Clemens analysis in field-service drive-time disputes.

Federal Skidmore application. A tight response SLA (5–15 min) + tracked acknowledgment + frequent calls pushes toward compensability. A long SLA (30+ min) + infrequent calls + ability to trade shifts pushes away from it. The California broader-than-FLSA analysis pushes further — Troester's no-de-minimis rule covers even brief acknowledgment moments.

Most tech employers pay an on-call stipend plus full hourly rate during active engagement. The open exposure is the standby time between calls — the time the employee can't meaningfully use because they're waiting for the next page.

Security and 24-hour guard services

The Mendiola pattern. 24-hour security at construction sites, residential facilities, event venues, and gated communities. The schedule is typically the Mendiola shape — active-duty hours plus on-call hours on the same premises. California exposure is the highest in the country post-Mendiola; other states apply Skidmore totality.

National security operators with uniform "we pay the active hours" policies face material California exposure even when the policy is defensible under federal law. The fix: either eliminate on-call from California schedules or pay the full 24 hours in California.

Field service emergency callbacks

The pattern overlaps with our travel time pay rules guide's §785.36 emergency-callback analysis. An HVAC tech, plumber, or IT support tech called out at 2am for an emergency customer repair earns:

• Compensable drive time to the emergency customer site under 29 CFR §785.36.
• Compensable on-call response time under §785.17 once they begin work.
• On-call standby time before the call — Skidmore totality decides whether the standby is also compensable.

Most field-service businesses pay actively-engaged callout time but treat the standby as uncompensated. The exposure surfaces when call logs are aggregated against the Skidmore factors and the standby starts looking restrictive.

Public safety — police and fire

Police and fire on-call patterns have their own statutory framework under FLSA §7(k) (29 USC §207(k) + 29 CFR Part 553), the public-safety exemption that allows a 7- to 28-day work period instead of the standard 7-day workweek. Most police/fire on-call disputes turn on §7(k) period definition rather than on-call doctrine itself. State law (CA, NY, NJ, IL) often gives police and firefighters collective bargaining rights that supersede the federal floor.

Multi-State and Remote Workers

On-call liability follows the employee's work location, not the employer's HQ. Same rule as travel-time, off-the-clock, overtime, and breaks. For multi-state employers:

• A Texas-headquartered tech company with a DevOps engineer in California on a PagerDuty rotation owes that engineer California "subject to control" analysis, not the federal Skidmore default. The same policy that's defensible in Texas may be a class action in California.
• A national hospital system with on-call physicians across multiple states applies each state's framework. Mendiola in California; Skidmore default in most other states; broader treatment in New York, Washington, and Oregon.
• A security operator with 24-hour-duty sites in multiple states needs per-state policies — California requires paying the full 24 hours; most other states permit the §785.22 sleep-time deduction with the four conditions in place.
• Post-pandemic remote workers with always-on Slack / email expectations face the same Skidmore-totality analysis. The "respond within 5 minutes to any DM" expectation is functionally on-call — and California's no-de-minimis rule (Troester) covers every brief acknowledgment moment.

The strict-everywhere defense: track on-call standby time in every state and pay it when Skidmore-compensable. A California-baseline on-call policy (all standby time captured, no de minimis, no §785.22 deduction) satisfies every other state's rule. The marginal labor cost is small; the elimination of per-state exposure is large.

Recordkeeping and the Mt. Clemens Rule

29 CFR §516.2 requires the employer to record all hours worked, including compensable on-call time. The Anderson v. Mt. Clemens Pottery burden-shifting rule applies: if on-call records are inadequate, the employee can establish unpaid hours by "just and reasonable inference," and the burden shifts to the employer to prove actual hours worked.

For on-call rotations, the practical implication is platform metadata. PagerDuty / Opsgenie / ServiceNow in IT; nursing-station handoff logs and EMR access records in healthcare; security-system event logs at construction sites. The same exact-time data that proves the employee responded to a 2am page also creates the record that the federal rule requires.

The defensive posture: capture the time the records will eventually surface anyway. See our recordkeeping requirements by state guide for the §516 retention windows. Where exact-time data exists in the platform metadata, the Camp v. Home Depot doctrinal chain (see our time clock rounding rules guide for the CA Supreme Court pending review) treats rounding away the recorded minutes as strict-liability exposure in California.

Connection to Overtime — the Regular Rate Trap

When on-call time IS compensable, it counts as "hours worked" for the FLSA overtime regular-rate calculation under 29 USC §207(a) — and three failure modes layer here, mirroring the travel time pay rules guide's regular-rate-trap analysis.

Failure mode 1: compensable on-call pushes the workweek over 40 hours

A non-exempt IT engineer works 35 billable hours and is on PagerDuty rotation for 168 hours (a full week). If Skidmore totality finds the standby time compensable (or 80 of the 168 hours compensable), the workweek crosses 40 — and FLSA §7 overtime owes 1.5× the regular rate on the excess. Most payroll systems compute overtime against billable hours only; the on-call hours are tracked separately for stipend purposes but never enter the overtime calculation.

Failure mode 2: on-call stipend in the regular rate

A weekly $200 on-call stipend is non-discretionary — announced in advance, paid as a matter of contract or past practice. Under 29 CFR §778.208 (non-discretionary bonus inclusion) and §778.209 (apportionment), the stipend must be included in the regular rate for overtime computation. The regular rate is the total straight-time compensation (including the stipend) divided by total hours worked.

Worked example. A technician works 40 billable hours at $30/hour ($1,200) plus a $200 weekly on-call stipend, plus 5 compensable on-call hours under Skidmore. Total compensation: $1,200 + $200 = $1,400. Total hours worked: 45. Regular rate (weighted average): $1,400 / 45 = $31.11/hr. Overtime owed: 5 hours × 0.5 × $31.11 = $77.78 (half-time premium on top of straight-time pay already received). A payroll system that pays only the stipend + the billable hours, without computing the half-time premium on the compensable on-call, owes $77.78 per workweek per technician — small per-employee, large at workforce scale and across the FLSA statute period.

Failure mode 3: two-rate compensation (standby + active)

When the employer pays a lower hourly rate during on-call standby ($10/hour) and the regular rate when actively engaged ($30/hour), 29 CFR §778.115 weighted-average regular-rate rule applies. The overtime regular rate is the weighted average, NOT the higher active-call rate. See our overtime rules by state Mistake #3 (bonus inclusion in regular rate) for the cluster-anchor coverage, and our how to calculate retro pay guide for the §778.303 retroactive-recompute mechanic when this gets corrected late.

The combined failure pattern: a technician on PagerDuty rotation paid hourly + on-call stipend + active-engagement premium, with no Skidmore analysis of the standby time. All three failure modes stack. An unwind requires recomputing the regular rate for every workweek in the statute period with on-call hours included, then doubling under §216(b) liquidated damages, then layering state penalty rules where applicable.

Recent Changes (2024–2026)

• Healthcare-sector wave (2024–2026). Hospital sleep-time + on-call class actions have surged as healthcare workers organize and as hospitals stretch staffing models. Most settlements aren't publicly disclosed but practitioner commentary tracks consistent six- to seven-figure outcomes for medium-sized hospital workforces.
• PagerDuty / on-call platform metadata as litigation evidence. Platform logs of acknowledgment, response, and resolution times have become routine evidence in IT/DevOps on-call disputes. The platforms create exact-time records that defeat estimation-based defenses, paralleling GPS records in field-service drive-time litigation.
• Post-pandemic always-on remote-work patterns. Customer-support, IT-help-desk, and inside-sales workers with tight response SLAs through Slack / email / Salesforce face Skidmore-style analysis when the standby time becomes the predominant constraint. California's no-de-minimis rule (Troester) compounds the exposure.
• DOL Wage and Hour Division guidance. No major regulatory amendments to §785.17 since promulgation; the DOL Field Operations Handbook continues to provide operational interpretation. DOL Fact Sheet #22 remains the canonical "hours worked" reference; state-specific guidance from CA DIR, NY DOL, and WA L&I has tightened in the same direction.

Frequently Asked Questions

If You Discover You've Been Doing This Wrong

On-call audits routinely uncover accumulated exposure: standby time that should have been compensable under Skidmore, sleep-time deductions that fail the §785.22 conditions in retrospect, on-call stipends that should have been included in the regular rate, California on-premises duty without the Mendiola adjustment. The unwinding playbook:

1. Audit by employee work location. Pull on-call rotation records, PagerDuty / Opsgenie / EMR access logs, call response times, and active-engagement timestamps for the full statute period (2 years federal, 3 if willful, 4 in California under unfair-competition law). For California employees, every minute of working-while-on-call counts under Troester.

2. Apply the Skidmore factors to each on-call schedule. Identify which standby periods are likely Skidmore-compensable. Response time, geographic limits, frequency of calls, personal-activity restrictions, ability to trade shifts. Document the analysis for each schedule type.

3. Pay back wages voluntarily. Self-correcting before a claim is filed is admissible as evidence of good faith. Federal recoveries include back pay plus liquidated damages (effectively double); California adds the §203 waiting-time penalty + unfair-competition treble damages exposure. Voluntary payment can eliminate the multipliers.

4. Restructure on-call schedules where exposure is concentrated. California 24-hour-duty patterns may need to be split into separate shifts. Tight-SLA tech rotations may need looser SLAs or true-handoff schedules. The structural fix often costs less than the back-pay tail.

5. Consult counsel for class-action exposure. Rough rule: more than 10 employees in California with consistent on-call patterns, or any systematic standby under-tracking across the workforce, crosses into class-action territory. Tyson Foods representative-evidence rule + Mt. Clemens burden-shifting make class certification dramatically easier when employer records are deficient.

The Through-Line

On-call exposure has three failure modes: paying a stipend without analyzing the underlying compensable hours (the most common pattern); failing the §785.22 sleep-time-deduction conditions on 24+ hour duty (especially in healthcare and security); and California 24-hour duty under the Mendiola rule (where the federal sleep-time safe harbor doesn't apply). Get all three right with good time records — platform logs + call response timestamps + mobile clock-in geolocation — and the Mt. Clemens burden-shifting rule works in your favor. Get any one wrong, and the back-pay math compounds across years and employees until the exposure dwarfs the underlying scheduling convenience.

For multi-state employers, the highest-leverage move is the same as for overtime, breaks, sick leave, and off-the-clock: standardize to the strictest applicable rule. California-baseline on-call capture (all standby time recorded, no §785.22 deduction, Skidmore analysis on every schedule) satisfies every other state's requirement. The marginal labor cost is small; the elimination of per-state policy complexity and the defensive posture against the Mt. Clemens burden shift are large.

Sources and Authorities

Federal

• 29 CFR §785.16 — Off duty (waiting time)
• 29 CFR §785.17 — On-call time
• 29 CFR §785.21 — Duty of less than 24 hours
• 29 CFR §785.22 — Duty of 24 hours or more (sleep-time deduction)
• 29 USC §207(j) — Hospital 8-and-80 alternative
• 29 CFR §778.601 — Hospital 8-and-80 implementation
• 29 USC §207(k) — Public-safety exemption (police and fire)
• 29 CFR §778.115 — Two regular rates (weighted average)
• 29 CFR §778.208 — Inclusion of bonuses in regular rate
• 29 CFR §778.209 — Method of inclusion of bonus
• DOL Fact Sheet #22: Hours Worked Under the FLSA

State

• California DIR — Hours Worked definition
• California IWC Wage Order 4 (Professional, Technical, Clerical)
• NY DOL — Labor Standards FAQ
• Texas Workforce Commission — Waiting or On-Call Time
• Washington L&I — Hours Worked

Case law

• Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680 (1946) — burden-shifting rule when employer recordkeeping is inadequate.
• Skidmore v. Swift & Co., 323 U.S. 134 (1944) — totality-of-circumstances test for on-call and waiting time.
• Armour & Co. v. Wantock, 323 U.S. 126 (1944) — companion case applying the same totality test to private firefighters.
• Berry v. County of Sonoma, 30 F.3d 1174 (9th Cir. 1994) — restricted off-premises on-call (30-min response + alcohol restriction + distance limit) NOT compensable under FLSA on these facts.
• Mendiola v. CPS Security Solutions, Inc., 60 Cal.4th 833, 340 P.3d 355, 182 Cal.Rptr.3d 124 (Cal. 2015) — on-premises 24-hour security-guard on-call fully compensable in California; federal §785.22 sleep-time deduction does not apply to California state-law claims.
• Troester v. Starbucks Corp., 5 Cal.5th 829 (2018) — California rejects federal de minimis doctrine.