Diwali 2026: Why the Five-Day Festival Runs Four Days, Not Five
Naraka Chaturdashi and Lakshmi Puja both fall on Sunday 8 November 2026, one date carrying two rituals while another date in the same week carries none. Neither is a calendar error. Here is the mechanism, called tithi kshaya, and what to actually do on 8 November, morning and evening.
Reviewed by Vidhata Editorial Desk · Updated
In this article
Search "Diwali 2026 date" and you will find almanacs that disagree with each other by a day, and a fair number of readers who conclude that one of them made an error. None of them did. In 2026 the five rituals of the Kartika sequence, Dhanteras, Naraka Chaturdashi, Diwali itself, Govardhan Puja and Bhai Dooj, land across six civil dates rather than five. One date, Sunday 8 November, carries two of the five rituals. Another date, the following Tuesday, carries none of them at all. That is what a lunar calendar running against a fixed count of solar days does periodically, and this year is one of the years it does it.
This piece is the mechanism, not the muhurat tables. For city-by-city bathing times, see Naraka Chaturdashi 2026: the pre-dawn bath, and why it shares a date with Diwali. For the pradosh and sthir lagna logic behind the evening puja, see Lakshmi Puja muhurat: pradosh and sthir lagna and Diwali and Lakshmi Pooja: the five-day festival explained. This one answers the question underneath both: why the dates fall where they do, and why that is nothing to worry about.
Five rituals, six dates
The published sequence for 2026 is Dhanteras on Friday 6 November, Naraka Chaturdashi and Diwali both on Sunday 8 November, Govardhan Puja on Monday 9 November, and Bhai Dooj on Wednesday 11 November. Laid out as a week:
| Date | Ritual |
|---|---|
| 6 Nov (Fri) | Dhanteras |
| 7 Nov (Sat) | nothing scheduled |
| 8 Nov (Sun) | Naraka Chaturdashi and Diwali (Lakshmi Puja) |
| 9 Nov (Mon) | Govardhan Puja |
| 10 Nov (Tue) | nothing scheduled |
| 11 Nov (Wed) | Bhai Dooj |
Five festival days spread across six calendar dates, with one date doing double duty and one date sitting empty. Both gaps come from the same root cause: a tithi is not twenty-four hours long, and its boundaries fall wherever the sky puts them, indifferent to which clock hour or which civil date is convenient.
What is actually happening on 8 November
A tithi is defined by the angular distance between the Moon and the Sun, not by the clock. Krishna chaturdashi, the tithi Naraka Chaturdashi needs, runs from 10:48 IST on 7 November to 11:28 IST on 8 November. The moment it ends, krishna amavasya begins, the tithi Diwali needs.
Lay the two rituals over that window. The abhyanga snan, the pre-dawn oil bath at the centre of Naraka Chaturdashi, is a sunrise-read observance: it is performed before the sun comes up, while chaturdashi is still running. Sunrise on 8 November falls between 05:45 and 06:49 depending on the city, and chaturdashi does not hand over to amavasya until 11:28, so every city in the country takes that bath comfortably inside chaturdashi.
The Lakshmi Puja, by contrast, is a pradosh rite: it is performed in the window after sunset, once amavasya is running and a fixed sign is on the ascendant. Amavasya has been in force since 11:28 that morning, so by the time pradosh arrives that evening, in Delhi from 17:30 to 20:08, in Mumbai from 18:01 to 20:33, amavasya has been running for hours. The puja belongs to that same 8 November.
So one civil day carries a dawn rite read off one tithi and a dusk rite read off the next. Both readings are correct in the instant they are taken. Nothing has gone wrong, and nobody's almanac is broken.
Tithi kshaya, properly
The stricter and more dramatic version of this same mechanism is worth explaining on its own terms, because it is the phrase most people reach for when they hear "a day got skipped" and it does not, strictly, describe what happened on 8 November this year.
A tithi is the time the Moon takes to gain twelve degrees of celestial longitude on the Sun. That interval is close to a day on average, but it is essentially never exactly a day, because the Moon's orbit around the Earth is an ellipse rather than a circle. Its angular speed against the background sky is not constant: it moves faster when it is nearer the Earth and slower when it is farther away, the same reason a planet on an elliptical orbit sweeps out equal areas in equal times rather than equal angles. The Sun's own apparent motion varies very slightly across the year for the same reason, though the Moon's variation is the larger effect. Because the rate at which the Moon closes each twelve-degree gap keeps changing, some tithis finish a little short of a full day and others stretch a little beyond one.
When a tithi runs short enough, something specific happens: it can begin after one sunrise and finish before the next sunrise arrives. In that case it never itself governs a sunrise. A panchang that assigns one tithi to each date by what is running at dawn has no date for it at all, because the date that would have belonged to it is never reached before the following tithi has already taken over. That is tithi kshaya, literally a decreased or lost tithi. It is not an error and nothing about the sky has actually skipped anything; the tithi occurred in full, it simply never touched a sunrise, so the sunrise-based count of dates passes over it.
That is a real and well-documented feature of Hindu lunar reckoning, and it is worth knowing on its own terms. It is also not, strictly, what produces 8 November's double booking. Both trayodashi and chaturdashi hold a sunrise this year, on 7 and 8 November respectively, so no tithi vanishes from the sunrise count in this stretch of the calendar. What compresses Naraka Chaturdashi and Diwali onto one date is a related effect of the same underlying variability, not a kshaya tithi itself: the chaturdashi-to-amavasya boundary happens to fall in the middle of the morning, comfortably clear of that day's sunrise and comfortably clear of that evening's pradosh, so both a sunrise-anchored rite and a pradosh-anchored rite can read their governing tithi off the same twenty-four hours. Different mechanism, same root cause: a tithi boundary is an astronomical instant that does not consult the ritual calendar before it falls.
The second gap: Bhai Dooj
The empty date on 10 November comes from the identical root cause working in the other direction. Bhai Dooj is an aparahna rite, performed in the fourth of the five equal divisions of daylight, and it needs shukla dwitiya running through that afternoon window. Dwitiya does not begin until 14:00 IST on 10 November, already inside the afternoon, so that day's aparahna window opens under the previous tithi and only picks up dwitiya partway through, too late to be the tithi the rite is built to use. By 11 November, dwitiya has been running since 14:00 the day before and covers that day's aparahna window outright, from 13:10 to 15:19 in Delhi. So the tilak waits a full day past Govardhan Puja, and 10 November passes with no Kartika-sequence ritual reading it as its own.
Nobody chose to leave a gap on the 10th any more than anyone chose to double up the 8th. Both are what happens when a count of tithis, each running close to but not exactly a day, is mapped onto a fixed run of solar dates.
What to do on 8 November 2026, in order
Before sunrise. Take the abhyanga snan while krishna chaturdashi is still running, which it is at every sunrise in the country this year. Sunrise falls between 05:45 in Kolkata and 06:49 in Ahmedabad, so set your alarm against your own city's time rather than a figure computed for somewhere else. City-by-city bathing windows are in the Naraka Chaturdashi post.
Through the day. Treat it as an ordinary Diwali eve: finish the cleaning, set up the Lakshmi-Ganesha altar, and have the evening's samagri ready before the light starts to go, since the pradosh window is not long and there is no benefit to starting the puja preparation late.
After sunset. Perform the Lakshmi Puja inside pradosh, once amavasya is running and a fixed sign sits on the ascendant. In Delhi that combined window runs 17:55 to 19:50; in Mumbai, 18:27 to 20:27. Full city-by-city windows, and the reasoning behind the fixed-sign requirement, are in the Lakshmi Puja muhurat post.
Neither rite has moved. Both belong to 8 November this year, one at its start and one near its end, and the ordinary sequence of the day carries you from one to the other without needing to treat either as displaced.
The reading behind the ritual
The calendar question and the personal one are different questions, and it is worth keeping them apart. Where the tithi boundaries fall this Kartika tells you when to perform which rite. What this stretch of the year is doing for you specifically depends on your own chart: which dasha and antardasha are running, and what the transiting planets, particularly the ones moving through Kartik, are doing against your natal Moon. A free kundali gives you that placement, and Ask Acharya can walk through what the current period means for you in particular, in your own language, with follow-up questions included at no cost.
The date confusion clears up once the mechanism is visible. The ritual reading of the year is the part that still needs your own chart.
Frequently asked
Common questions
Why do Naraka Chaturdashi and Diwali fall on the same date in 2026?+
Because the tithi that governs the pre-dawn abhyanga snan, krishna chaturdashi, is still in force at sunrise on 8 November, and the amavasya that governs the Lakshmi Puja begins later that same morning and is running by the time pradosh, the window after sunset, arrives. Sunrise reads one tithi and pradosh reads the next, and this year both windows fall inside the same civil day. Both readings are correct at the moment they are taken.
What is tithi kshaya?+
A tithi is the time the Moon takes to gain twelve degrees on the Sun, and that duration is not fixed at twenty-four hours because the Moon's orbital speed varies through the month. When a tithi runs short enough that it begins after one sunrise and finishes before the next, it never itself holds a sunrise, so a strict sunrise-based panchang has no date for it at all. That lost date is tithi kshaya, a decreased or vanished tithi. It is a routine feature of a lunar calendar laid over solar days, not a mistake anyone made.
Is what's happening on 8 November 2026 actually tithi kshaya?+
Not in the strict sense. Both trayodashi and chaturdashi hold their own sunrise this year, on 7 and 8 November respectively, so no tithi disappears from the sunrise reckoning. What compresses the festival is a related but distinct effect of the same variability: the chaturdashi-to-amavasya boundary falls mid-morning on the 8th, after that day's sunrise and well before that evening's pradosh, so the dawn rite and the dusk rite end up reading their tithi off the same day.
Why is Bhai Dooj on 11 November and not 10 November?+
Shukla dwitiya, the tithi Bhai Dooj needs running through the aparahna window, does not begin until 14:00 on 10 November, already inside the afternoon. By 11 November it has been running since the previous afternoon and covers that day's aparahna window outright. The tilak is an aparahna rite, not a sunrise rite, so it waits for the day the tithi actually covers that window, which pushes it a full day past Govardhan Puja.
What should I do on 8 November 2026, and in what order?+
Take the abhyanga snan, the ritual oil bath, before your own sunrise, since that is when krishna chaturdashi governs the morning. Go through the day as an ordinary Diwali eve, cleaning and preparing the altar. In the evening, perform Lakshmi Puja inside pradosh, the window after sunset, once amavasya is running and a fixed sign is on the ascendant. Both rites belong to 8 November; neither has moved to a different date.
Are these times the same everywhere in the world?+
The tithi boundaries themselves are the same instant everywhere on Earth. What changes by location is only the local clock reading of sunrise, sunset and pradosh, which decides which civil date each boundary falls inside. A household abroad should check its own sunrise and pradosh rather than borrow a time computed for an Indian city.