Shastra Panchangam

Why ekadashi falls on two different days

Because the two traditions test the tithi at different moments — Smartas at sunrise, Vaishnavas four ghatikas before it.

You will sometimes see two ekadashi dates printed for what is obviously the same ekadashi, one of them marked Vaishnava. That is not a mistake and it is not a regional variant. It is two decision rules, both written down, producing different answers.

Vedha

The word means piercing. An ekadashi is viddha — pierced — when the previous tithi, dashami, is still running at the moment the tradition tests for it. A pierced ekadashi is not fasted on.

Everything turns on when you test.

The two tests

Smarta. The test is at sunrise. If dashami is still running when the Sun comes up, the ekadashi is pierced.

Vaishnava. The test is at arunodaya — four ghatikas before sunrise. A ghatika is 24 minutes, so this is 96 minutes earlier, in the dark. If dashami extends even a kala past the 56th ghatika of the day, the ekadashi is pierced for Vaishnavas.

The Vaishnava test is the stricter of the two, because it is applied earlier and dashami is more likely to still be running. So a pierced ekadashi is abandoned and the fast moves to dvadashi, the following day.

What that produces

On most ekadashis both tests agree and one date is printed. When dashami happens to end inside that 96-minute window — after arunodaya but before sunrise — the tests disagree, the Smarta fast is on one day and the Vaishnava fast on the next, and the almanac prints both.

The established almanacs split on 23 and 24 August 2026 — the first plain, the second carrying the Vaishnava mark. That is the rule above doing its work, not a disagreement between publishers.

Which one applies to you

Your household's, and this page does not yet tell you which date that is. The Smarta day follows from the tithi at sunrise and is unambiguous. The Vaishnava day does not follow from the arunodaya test alone: on 23 August 2026 the ekadashi is suddha, and the established almanacs still mark the 24th, which comes from a dvadashi persisting past the following sunrise. That is one of the eight adhikya cases, and we hold only a truncated statement of them.

So we publish the rule and not the list. A Smarta-only list would read as the ekadashi date to half the households that found this page, and a wrong observance date is worse than an absent one.

Source
Dharmasindhu (Kāśīnātha Upādhyāya, 1790), pariccheda 1, athaikādaśīvratadinanirṇayaḥ and atha smārtanirṇayaḥ, pp. 13–14, leaves 33–34 · read from the page image
with Swamikannu Pillai, Indian Chronology