Diwali 2027 Date and Time: Puja Windows for 18 Indian Cities

Diwali 2027 Date and Time: Puja Windows for 18 Indian Cities

The Diwali 2027 date and time: Lakshmi Puja falls on Friday 29 October 2027, and the evening puja runs from your own city's sunset until Amavasya ends at 7:06 PM IST, one instant for the whole country. That leaves Guwahati 142 minutes of Amavasya after sunset and Mumbai only 59, and the table below works the window out for 18 cities.

If you also plan to send a greeting, expect to make it on the day. In SubhSandesh's Raksha Bandhan 2026 data, our nearest festival analogue (Rakhi, not Diwali, and only n = 89), 55.7% of the pages finished in time were built on the festival day itself, and across the platform 48.3% of shared pages are opened on a phone. Below: the city table, why the east and west differ by more than an hour, why Thursday never qualifies, and how every minute was computed.

What "date and time" means for Diwali 2027

The date comes from the tithi. By our calculation Amavasya begins at 8:49 PM IST on Thursday 28 October and ends at the new moon, which the US Naval Observatory puts at 7:06 PM IST on Friday 29 October. Lakshmi Puja is done in pradosh, the stretch just after sunset, on the evening Amavasya covers it, and that evening is Friday.

The Report of the Calendar Reform Committee (1955), prepared for the Government of India, records pradosh as two muhurtas after sunset and notes that some authorities take three. The same report defines a night muhurta as one-fifteenth of the night, not a flat 48 minutes. On 29 October 2027 that makes one night muhurta 48.8 minutes long in Kochi and 51.9 minutes in Chandigarh, because the northern night is longer.

Tamil Nadu is the known exception on the date. Tamil Deepavali follows Naraka Chaturdashi at dawn, and drik calculations put it on Thursday 28 October, while some holiday listings show the 29th; check your state's 2027 holiday notification before planning around either.

Diwali 2027 time in 18 cities, east to west

The table lists each city's sunset on Friday 29 October 2027, the length of one night muhurta there, and the puja window on the 1955 report's two-muhurta reading, cut off at 7:06 PM. Cities run from the earliest sunset to the latest.

Every city west of Varanasi loses its window to the 7:06 PM cut-off, and the further west, the shorter it gets, down to 59 minutes in Mumbai. Your panchang will narrow it further to a fixed lagna, which we do not print.

Why Guwahati gets 142 minutes of Amavasya and Mumbai only 59

India runs on one clock set to the meridian at 82°33′E, yet the country spans from 68°7′E to 97°25′E, roughly two time zones' worth. A 2018 Current Science paper by the CSIR-National Physical Laboratory, the custodian of Indian Standard Time, found the sunrise gap between Dong in the far east and Ghuar Mota in the far west close to two hours at the equinox.

Guwahati (91.74°E) and Mumbai (72.88°E) are 18.9° apart, worth about 75 minutes of solar time, but their sunsets on 29 October 2027 differ by 83 minutes. Latitude supplies the rest: the same paper notes that lines of equal sunrise follow longitude only near the equinoxes, and late October is five weeks past one.

Chennai shows the latitude effect best. It sits about 3° east of Delhi, which should bring its sunset roughly 12 minutes earlier, yet the sun sets there at 5:43 PM against Delhi's 5:39 PM. Further south the late-October day is longer, so Chennai's night lasts 739 minutes against Delhi's 773.

Five eastern cities where pradosh closes before Amavasya

On the two-muhurta reading, pradosh ends before 7:06 PM in five cities: Guwahati at 6:26, Kolkata at 6:42, Patna at 6:53, Bhubaneswar at 6:54 and Varanasi at 7:02 PM. There the window is sunset plus about 100 minutes, with Amavasya still running when it closes.

Lucknow is the city where the definition decides the answer. With a flat 48-minute muhurta its pradosh would end at 7:02 PM, before Amavasya does; with the 1955 report's one-fifteenth-of-the-night muhurta of 51.2 minutes, it ends at 7:08 PM, and Amavasya closes the window at 7:06.

On the three-muhurta reading that some authorities allow, Amavasya is the limit in all 18 cities. Even Guwahati's longer pradosh then runs to 7:17 PM, 11 minutes after the tithi has ended, so an eastern family's window becomes sunset to 7:06 PM: 142 minutes in Guwahati and 125 in Kolkata.

Why no Indian city does Lakshmi Puja on Thursday 28 October 2027

Amavasya begins at 8:49 PM IST on Thursday, by our calculation, and some panchangs print 8:47 PM. Even on the generous three-muhurta reading, Thursday's pradosh is over by then in every city in the table: it ends at 7:17 PM in Guwahati, 8:14 PM in Delhi and 8:37 PM in Mumbai.

Mumbai comes closest, missing by 12 minutes, so no Indian city has Amavasya inside Thursday's pradosh and the puja is Friday everywhere in India. Outside India the answer can change, because the 7:06 PM IST cut-off falls before sunset in the Gulf, Europe and the Americas.

Why apps differ by a minute or two, and how we computed ours

We computed every sunset with Astronomy Engine 2.1.19, an open-source ephemeris library, at city-centre coordinates. We then checked it against the US Naval Observatory's one-day service, which gave the same minute for Guwahati, sunset 4:44 PM and Ahmedabad, sunset 6:04 PM. Both responses list the new moon at 7:06 PM IST; Astronomy Engine puts it one minute later, so we used the observatory's figure.

Three things move the minutes between apps. The sunset convention is one: the observatory times the sun's upper edge, while the 1955 report timed its centre with 30′ of refraction, which puts Delhi's sunset 2 minutes earlier at 5:37 PM. Coordinates matter less than people expect: Navi Mumbai sets one minute before central Mumbai. The third is rounding of the Amavasya end, which online panchangs print between 7:05 and 7:07 PM.

None of this touches the lagna. A fixed lagna such as Vrishabha rises at a different IST minute in every city, and we have no non-competitor almanac to source it from, so read it off a panchang set to your own city.

Sending a Diwali greeting around your city's puja window

Of the 79 Raksha Bandhan 2026 pages finished in time, 86.1% were built within 24 hours of the festival (again Rakhi data, n = 89), and a shared page averages 11.1 views, so it gets opened more than once.

The windows also tell you who is free when. A sister in Kolkata whose two-muhurta window closes at 6:42 PM is free again while a brother in Mumbai is still inside his 6:07 to 7:06 PM window, so a greeting sent at 6:50 PM suits her and interrupts him. For that one person, the interactive Happy Diwali card on SubhSandesh lets them light a diya waiting at their door with your tag on it, light a row of wish diyas, colour a rangoli until your initials glow and write their name with a sparkler.

It holds up to six photos in sky lanterns, seals your letter in a shagun liffafa and plays a YouTube song or a Diwali favourite. Setup takes about five minutes in love, family or friend mode, and the private link opens in any browser with no app or account.

For a parent in another city who would rather hear you than tap through scenes, a voice dedication that plays like a radio show fits better. It is not a Diwali template, which is its limit: there are no diyas, only your voice introducing a song.

When this timetable is the wrong tool

It gives no lagna minutes, so it cannot replace your panchang's Vrishabha or Simha lagna table. It also says nothing about Kali Puja, which Bengal times to midnight under a different rule, or about towns far from a listed city: each degree of longitude, roughly 100 km in central India, moves sunset by about 4 minutes, so compute your own.

Your family's priest and tradition outrank every number here, including the 1955 report, which recorded many regional practices rather than one correct rule. And for a grandparent who does not use a smartphone, a link is the wrong gift: 48.3% of our shared pages open on a phone, and a phone call after their puja still beats both.

Sources

• US Naval Observatory, Astronomical Applications one-day data for Guwahati and Ahmedabad, 29 October 2027 (computed service, no publication date; queried 22 October 2026).
• Calendar Reform Committee, Report of the Calendar Reform Committee, CSIR, Government of India, 1955, Internet Archive.
• Lakhi Sharma and others, "Necessity of two time zones: IST-I (UTC+5:30 h) and IST-II (UTC+6:30 h) in India and its implementation", Current Science 115(7), 10 October 2018, pp. 1252 to 1261.
• Don Cross, Astronomy Engine 2.1.19 (MIT licence), used for all 18 cities' sunrise, sunset and tithi times.