The Four Day Work Week in India: What the Labour Codes Allow, and Why Four Twelve Hour Days Do Not Fit

Since the final Central Rules under India's four Labour Codes were notified on 8 May 2026, one line has travelled further than any other: you can now run a four day work week. Four days on, three days off, twelve hours a day, forty eight hours a week. The arithmetic is clean, it fits in a headline, and almost every write-up published since May has repeated it.

The arithmetic is also wrong. Not the multiplication, which is fine, but the assumption underneath it: that the 48 hour weekly cap is the only number governing how you lay out a day. It is not. There are two more, they sit in the same framework, and together they make a twelve hour working day impossible to roster lawfully in the central sphere. The most work that fits in a compliant day is eleven hours, which makes the real compressed week 44 hours, not 48.

This guide is the version that does the full sum. It covers what the Codes actually permit, the two provisions that break the popular version, the roster shapes that are genuinely lawful, why you cannot impose any of them unilaterally, what happens to overtime and weekly offs, and the records you will be asked for. It assumes you have read nothing else. If you want the wider compliance picture around it, our 12 point Labour Code compliance checklist for 2026 covers the other eleven obligations, and our guide to overtime calculation covers the formulas this piece assumes.

What the Labour Codes Actually Say About a Four Day Work Week

Start with what is genuinely new, because it is real and it is useful. Nothing in the Occupational Safety, Health and Working Conditions Code, 2020 requires the 48 weekly hours to be spread across six days. The Code fixes a weekly ceiling and a daily ceiling and leaves the distribution to the establishment. That is the flexibility everyone is reacting to, and it is a genuine change from a Factories Act regime built around a six day pattern.

Three numbers govern every roster you can build. Two of them are quoted constantly. The third almost never is.

The three numbers that govern every roster

Forty eight hours a week. This is the hard ceiling on ordinary working hours. It is not negotiable, it is not averageable across a month, and work beyond it is overtime at twice the ordinary rate of wages. Distributing it across four, five or six days is permitted. Exceeding it is not.

Eight hours a day, extendable. Section 25 of the OSH Code sets the ordinary working day at eight hours. Hours beyond eight in a day are overtime hours and are paid at 2x, even in a week that never reaches 48. This is the point most four day week discussions skip: a twelve hour day is not four extra ordinary hours, it is four overtime hours, and they cost double.

Twelve hours of spread-over. Spread-over is the total elapsed time from the first clock-in to the last clock-out on a day, including unpaid breaks. The ceiling was raised from 10.5 hours to 12 hours under the new framework, which was widely reported as an enabler of the long day. It is the provision that actually forbids it.

The Arithmetic Problem: Four Twelves Do Not Fit

On 13 May 2026, five days after the Central Rules were notified, the Ministry of Labour and Employment issued a notification under Section 25(1)(b) of the OSH Code fixing the rest interval. No worker may work more than five hours continuously without a break of at least thirty minutes.

That notification is the one nobody put next to the spread-over cap. Do it now.

Why twelve hours of work needs thirteen hours of clock

To work twelve hours without ever running past five hours continuously, you need at least two rest intervals. The shortest lawful arrangement is five hours of work, thirty minutes of break, five hours of work, thirty minutes of break, then the remaining two hours. Add it up:

5 + 0.5 + 5 + 0.5 + 2 = 13 hours of elapsed clock for 12 hours of work.

Thirteen hours of spread-over against a ceiling of twelve. A worker who clocks in at 8 AM finishes at 9 PM. There is no way to compress it: the breaks are a floor, not a target, and shortening them below thirty minutes or stretching continuous work past five hours breaches the 13 May notification instead.

What actually fits inside a twelve hour spread-over

Run the same arrangement backwards from the ceiling. Twelve hours of elapsed clock, minus the two mandatory thirty minute intervals, leaves eleven hours of work: five hours, break, five hours, break, one hour. Every segment is inside five continuous hours and the whole day closes exactly on the twelve hour spread-over line.

Eleven hours is therefore the true daily maximum in the central sphere, not twelve. Which makes the honest four day week:

4 days x 11 hours = 44 hours. Four hours short of the 48 you are entitled to roster.

A caveat worth stating plainly, because this is the kind of claim that ought to carry one. No authority has yet been asked to reconcile the spread-over ceiling with the 13 May rest interval notification, and no clarification has been issued. The reading above is the arithmetic that follows from the two provisions as published. If a clarification arrives that relaxes the spread-over for compressed schedules, the twelve hour day becomes available and this section becomes history. Until then, rostering a twelve hour day and assuming it is covered by "4 x 12 = 48" is a risk you are taking without knowing you took it.

The Compressed Rosters That Are Actually Lawful

Here is what remains once the arithmetic is honest. Each of these sits inside the 48 hour cap, the 12 hour spread-over and the five hour break rule.

Four days of eleven hours. 44 hours. The genuine three day weekend. You give up four hours of ordinary capacity per person per week, and twelve of the 44 hours are overtime hours at 2x, because everything past eight in a day is. Cheapest in calendar terms, most expensive per hour.

Four long days plus one short day. 48 hours. Four days of eleven hours and a fifth day of four. You keep the full 48 and still hand back a genuinely lighter week. In practice this is the shape most Indian SMEs will land on once someone does the sum.

Five days of nine hours and thirty six minutes. 48 hours.Unglamorous and fully compliant, with only one overtime hour and thirty six minutes per day above the eight hour trigger. For a business that cannot lose a fifth day of coverage, this is the version of "flexibility" the Codes actually deliver.

Six days of eight hours. 48 hours. The status quo, and still lawful. Nothing in the Codes obliges you to change anything.

Notice what happens to cost across that list. The 48 hour cap is constant, but the number of hours sitting above the eight hour daily trigger is not, and those hours are priced at double. A four day, eleven hour roster pays 2x on twelve hours a week per worker. A six day, eight hour roster pays 2x on nothing. The three day weekend is not free, and the invoice arrives in the overtime line.

You Cannot Impose It: Consent Is Structural, Not Courtesy

The compressed week is permissive, not mandatory, and the permission runs in only one direction. An employer may offer it. An employer may not impose it. Legal commentary on the Central Rules has been consistent on this: a four day week requires the agreement of the workforce and is not an automatic entitlement on either side.

Two consequences an SME owner should act on rather than note:

Consent has to exist as a record, not as a conversation. If you roster an eleven hour day and a worker later says they never agreed to it, the question in front of an inspector is what you can produce. A signed variation to the appointment letter, which the Codes now require every worker to hold anyway, is the cleanest place to put it. A WhatsApp thread is not a record of consent.

Overtime carries its own consent, separately.Overtime under the OSH Code requires the worker's consent and refusal is not a disciplinary matter. Since a compressed roster puts three or four overtime hours into every single working day by design, you are asking for that consent daily, structurally, for the life of the roster. Agreeing to a four day week and agreeing to work overtime on demand are not the same agreement, and it is worth writing them down as two.

Weekly Off Rules Under the Labour Codes in 2026

A compressed week changes which days are rest days, so the weekly off rules stop being background and start being operational.

One rest day a week, at least twenty four consecutive hours. Every worker is entitled to a weekly day of rest. Under the Central Rules it ordinarily falls on Sunday for a six day week, and on Saturday and Sunday for a five day week.

You may substitute the rest day, with notice. Employers can designate an alternative rest day, provided workers are told in advance through a formal communication rather than a shout across the floor. On a four day roster the two extra rest days are not statutory weekly offs, they are simply days you have not rostered, and that distinction matters when you want someone to come in.

Work on a rest day earns a substituted day of rest.Not cash instead, by default. If a worker comes in on their designated weekly off, they are owed a compensatory off. Treating the compressed week's spare days as a bank of callable overtime days is exactly the pattern that turns a popular roster into a dispute.

The ten consecutive day limit is not from the Codes. The often quoted rule that a worker cannot go more than ten consecutive days without a holiday comes from the substitution provision of the Factories Act, and it bounds how far you may push a rest day when you substitute one. It is being repeated as though it were a new Code provision. It is not, and if you sit in the state sphere under a Shops Act rather than the Factories Act, check what your own statute says instead of importing this number.

The Split Shift Trap: Spread-Over Catches More Than the Four Day Week

The twelve hour spread-over cap is being read as a rule about long shifts. It is not. It is a rule about long days, and it catches a group of employers who have never considered themselves anywhere near a twelve hour roster: anyone running a split shift. Restaurants, clinics, salons, tuition centres and retail with a dead afternoon all do this, and most of them do it in a shape that does not survive the cap.

The gap is not a break, and it is not free. This is the point that catches people. Unpaid time between the two halves of a split shift still sits inside the spread-over, because spread-over measures elapsed time from first arrival to final departure, not time on the clock. An employer who reasons that the worker went home, so it does not count, has the arithmetic exactly backwards.

It is invisible in most attendance records. A register that marks a person present cannot produce a spread-over figure at all, and a system that stores only a daily total will report ten hours and look fine. The breach only appears if you hold the first and last timestamps of the day and subtract one from the other, per person, per day.

Night Shifts and Women Workers on a Compressed Roster

A compressed roster pushes shift ends later almost by definition, which brings a second set of rules into play. These are a genuine liberalisation, and they arrive with conditions attached that are the employer's to provide rather than the worker's to waive.

Women may work the night window, across all sectors. The blanket prohibition on employing women before six in the morning and after seven in the evening is gone under the OSH Code. What replaces it is a consent based regime, and the consent is specific: it must be in writing, it must be voluntary, and it can be withdrawn. Consent taken as a condition of being hired is not consent.

The safeguards are obligations, not recommendations. Where women are employed in the night window the employer has to provide transport to and from the residence, adequate lighting across the premises and the approach to them, security personnel and surveillance through the shift, and a grievance mechanism with women represented on it. An employer who obtains the consent and skips the transport has not complied with anything.

Several states have layered their own conditions on top. Shops and commercial establishments in particular are covered by state notifications that add requirements, and those apply to you before any central rule does if you sit in the state sphere. Check the notification that governs your establishment before you roster a night shift rather than after somebody has worked one.

A shift crossing midnight moves the weekly off. The twenty four consecutive hours of weekly rest are counted from the end of the shift, not from midnight. On a late finishing compressed roster that quietly shifts when the rest day actually begins, and a rest day recorded against the wrong calendar date is the kind of error that only surfaces when somebody counts back through the register.

Which Overtime Cap Binds You: Central Sphere or State Sphere

Because a compressed roster generates overtime hours every day, the quarterly overtime cap stops being theoretical and becomes the constraint that decides whether the roster survives a full quarter.

The OSH (Central) Rules, 2026 set the quarterly cap at 144 hours. That figure has been reported everywhere as "the new limit" with the second half of the sentence removed. It is the limit for central-sphere establishments: mines, major ports, railways, oilfields, air transport, banking, insurance, central public sector undertakings, and establishments operating in more than one state.

Almost every Indian SME is in the state sphere. The shop, the restaurant, the clinic, the single state factory, the gym, the logistics yard. If your state has not yet notified its own OSH rules, your existing state cap continues to bind you, and those run from around 50 hours a quarter in several states to 125 in others. Maharashtra, Gujarat, Karnataka, Madhya Pradesh and Delhi had notified rules across all four Codes as of mid 2026. Many states had not.

Now put that against a four day, eleven hour roster. Twelve overtime hours a week, roughly thirteen weeks a quarter, is about 156 overtime hours per worker per quarter. That is past the central cap of 144 and roughly triple a 50 hour state cap. A four day, eleven hour week run continuously is not a scheduling choice at all in most states. It is a cap breach with a three day weekend attached.

This is the sentence the popular version leaves out entirely, and it is the one that matters most: the roster is lawful for a few weeks and unlawful across a quarter. Which is why the four long days plus one short day shape, or a compressed roster used seasonally rather than permanently, is usually the only version that survives contact with the cap.

What a Compressed Week Does to Your Payroll

The overtime formula does not change. Hours past eight in a day, or past 48 in a week, are paid at twice the ordinary rate of wages, computed on the statutory wage base:

Overtime pay = ((Basic + DA) ÷ (26 × 8)) × 2 × OT hours

Worked example. A worker with Basic + DA of Rs 20,800 a month moves to a four day, eleven hour roster. The hourly ordinary rate is 20,800 ÷ 208, which is Rs 100. Three overtime hours a day at 2x is Rs 600 a day, Rs 2,400 a week, roughly Rs 10,400 a month in overtimeon a Rs 20,800 wage. The compressed roster has added fifty percent to that worker's monthly cost while reducing weekly ordinary capacity from 48 hours to 44.

That is not an argument against a four day week. Retention, absenteeism and recruitment can easily justify it. It is an argument against adopting one because a headline said the hours were the same. They are not the same hours. They are the same total with a different price on the last three of each day. Our free overtime calculator will run your own figures, including the divisor question, with no signup.

One further trap specific to 2026. The wage base these calculations sit on has itself changed. Under the Code on Wages, if excluded allowances exceed fifty percent of total remuneration the excess is added back into wages, which raises Basic + DA for overtime purposes as well as for PF and gratuity. If you restructured salary this year and did not re-point the overtime formula at the new base, every overtime hour on the compressed roster is being underpaid. Our guide to the new salary structure covers that add-back in full.

The Records an Inspector Will Ask For

A compressed roster multiplies the number of things you have to be able to prove. Six of them, and every one is derived from attendance data rather than from a policy document:

Actual hours worked per worker per day, not scheduled hours. A roster is an intention. The register has to show what happened.

Spread-over per day. First clock-in to last clock-out. This is the number that decides whether the eleven hour day stayed lawful, and almost no Indian SME records it today because under a nine hour day it never mattered.

Rest intervals taken. Not merely scheduled. The five hour rule is about continuous work, so a break that was rostered and not taken is a breach that a rostered break sheet will not reveal.

Overtime hours, per worker, running to a quarterly total. Against the cap that actually binds you, which as above is probably not 144.

Weekly offs taken, and compensatory offs owed where a rest day was worked.

Written consent to the compressed schedule, and to overtime.

The Central Rules require an Attendance Register cum Muster Roll, a Register of Wages, and an Overtime register among others. All three are computed from the same underlying record. Our guide to statutory registers and wage slips covers the full set and what each one has to contain.

Running a Compressed Roster Without a Spreadsheet Argument

Every item on that list resolves to one capability: knowing, provably, who worked which hours, with which breaks, at which location. A roster in a spreadsheet cannot produce any of it, because a spreadsheet records what you intended rather than what occurred. This is the specific gap Shiftelio was built to close, and a compressed week is close to the hardest test of it.

Shift patterns with an eight hour overtime trigger. Ordinary hours are set on the shift pattern itself, so an eleven hour compressed day flags three overtime hours automatically and prices them at 2x on Basic + DA, with the divisor configurable for your state.

Spread-over shown per day, per worker. First punch to last punch, so a day that drifted from eleven hours to twelve and a half is visible the same evening rather than in a quarterly audit.

Running quarterly overtime totals with a warning before the cap. Since a compressed roster reaches a 50 hour state cap in roughly four weeks, a running total is the difference between steering around the limit and discovering it.

Verified punches, not honour system punches. Each punch is a live selfie with GPS geofencing and mock location detection, on the phones staff already carry, with no biometric machine at the door. On an eleven hour day the difference between rostered and actual hours is where the entire compliance question lives.

Registers generated from the same record. The attendance register, wage register and overtime register are computed from one set of numbers, so they cannot disagree with each other. You can see the shape of the output in our free shift roster template and attendance register generator, both usable without an account.

Frequently Asked Questions

Is the four day work week now legal in India?

A compressed week is permitted, in the sense that nothing requires 48 hours to be spread over six days. But a four day week of twelve hour days does not fit inside the 12 hour spread-over ceiling once the mandatory rest intervals are counted. Four days of eleven hours is 44 hours and is lawful. Four days of twelve hours is not, on the plain reading of the current provisions.

Can my employer force me onto a four day, twelve hour roster?

No. The arrangement is permissive and requires agreement. It cannot be imposed unilaterally, and overtime, which a long day generates every day, separately requires the worker's consent, with refusal not being a disciplinary matter.

Does a four day week mean less pay?

It should mean more. Ordinary weekly hours fall from 48 to 44 on a four day, eleven hour roster, but three hours of every working day sit above the eight hour daily trigger and are paid at twice the ordinary rate. On a Rs 20,800 Basic + DA that is roughly Rs 10,400 a month in overtime.

What is spread-over and why does it matter now?

Spread-over is total elapsed time from first clock-in to last clock-out, including unpaid breaks. The ceiling is 12 hours. Under a nine hour day it was never close and nobody tracked it. Under an eleven hour day it is the binding constraint, and it is measured on actual punches rather than on the roster.

How many overtime hours can I roster in a quarter?

144 hours if you are a central-sphere establishment. If you are in the state sphere, which covers almost every Indian SME, your existing state cap applies until your state notifies its own OSH rules, and those commonly run from 50 to 125 hours. A four day, eleven hour roster generates roughly 156 overtime hours a quarter, which breaches every one of those figures.

Do the extra days off count as weekly offs?

One of them does. The statutory entitlement is one rest day of at least 24 consecutive hours per week, and working it earns a compensatory day of rest. The additional non-working days on a compressed roster are simply days you did not roster, so they are not protected in the same way.

The Bottom Line

The four day work week in India is real, it is permitted, and it is not the roster that has been described in print since May. Forty eight hours across four days requires twelve hour days, twelve hour days require thirteen hours of clock once the mandatory breaks are counted, and thirteen hours breaches the spread-over ceiling. What fits is eleven hours a day and 44 hours a week, or four long days plus one short one if you want to keep all 48.

Both of those shapes put three overtime hours into every working day, at double rate, running toward a quarterly cap that most Indian SMEs will hit inside a month. That is not a reason to reject the idea. It is a reason to price it before announcing it, and to run it on a record that shows actual hours, actual breaks and actual spread-over rather than the roster you meant to run.

Every business already generates that data. The only question, as with almost everything in the Codes, is whether it lives somewhere it can be read back at the end of a quarter, or in a roster sheet that was last accurate on the Monday it was printed.

Sources and Further Reading

The primary material for this article is the Occupational Safety, Health and Working Conditions Code, 2020 and the Central Rules notified under it on 8 May 2026. The Press Information Bureau publishes an official factsheet on the OSH Code covering working hours, overtime and welfare provisions, and the full statutory text is on the Government of India's India Code repository. Notification status for the Codes and the state rules is tracked by the Ministry of Labour and Employment.

This article describes the position as at 4 September 2026. State rules are still being notified through the year, and the interaction between the spread-over ceiling and the rest interval notification has not yet been the subject of any clarification. It is general information, not legal advice. Confirm your own state's notified position before changing a roster.