The 99 names of Allah

What the sahih hadith actually says

Two of the strongest collections carry this from Abu Hurayrah, and they agree. Allah has 99 names, one hundred less one; whoever does a certain thing with them enters Paradise; and Allah is witr and loves witr.

That's the whole hadith. There's no list in Bukhari 6410 and no list in Muslim 2677. The number is established. Which 99 they are is simply not stated.

• Sahih al-Bukhari 6410 (Book 80, Hadith 105), narrated by Abu Hurayrah.
• Sahih Muslim 2677 (Book 48, Hadith 5), same narrator, same chain through al-A'raj and Abu az-Zinad.

The list is in a different hadith, and it's weak

The 99 names as a list appear in Jami' at-Tirmidhi 3507. Sunnah.com prints the Darussalam grading for it as da'if, which is weak.

That single fact settles most of the argument, but it isn't the strongest thing available. The strongest thing is what Tirmidhi wrote directly underneath his own narration.

Tirmidhi said the chain isn't sound, in his own book

After recording the hadith with the names, Abu Isa at-Tirmidhi adds a comment. He calls it gharib. He says he knows it only through Safwan ibn Salih, who he notes is trustworthy among the hadith scholars. Then he says the part that matters: across the various narrations of this hadith, he doesn't know of a sound chain that mentions the names, except in this one. And where Adam ibn Abi Iyas narrated it by a different chain that also listed the names, he says plainly that it has no sound chain.

So the compiler of the collection everyone quotes for the list is on record, in the same book, saying the list has no sound chain behind it. That isn't a modern critique or a minority opinion. It's printed alongside the text.

There are two lists, and they don't match

This is the part that usually gets left out. Tirmidhi isn't the only collection with a list. Sunan Ibn Majah 3861 has one too, also graded da'if, and it's a different list.

We compared them. Matching the Arabic after stripping the vowel marks, and treating 5 pairs as one name where they're the same name in a different grammatical form (al-Muta'ali and al-Muta'al, al-Barr and al-Bar, ar-Rashid and ar-Rashid, al-Hafiz in its two forms, and al-Qahhar and al-Qahir), 77 names appear in both lists. Around 21 in Tirmidhi's list are missing from Ibn Majah's, and around 22 in Ibn Majah's are missing from Tirmidhi's.

Roughly a fifth of each list is absent from the other. If either list were the Prophet's own enumeration, the two would agree. Names that sit in one and not the other include al-Muqsit, al-Muhsi and Dhul-Jalali wal-Ikram on Tirmidhi's side, and al-Qarib, al-Jamil and al-Kafi on Ibn Majah's.

The list, as Tirmidhi records it

Here it is as printed, so you can read what's actually narrated rather than a version that's been tidied up somewhere along the way. It runs from ar-Rahman to as-Sabur and comes to 99 names.

What 'whoever ahsaha' means

The reward clause is worth its own line because the wording differs between the two sahih narrations, and the difference isn't small.

Muslim's main wording is man hafizaha, whoever commits them to memory. In the narration of Ibn Abi Umar the wording is man ahsaha, whoever enumerates or reckons them. Bukhari's wording has la yahfazuha ahad illa dakhal al-jannah.

Scholars have read ahsaha in more than one way: counting them, memorising them, understanding what they mean, or living by them. The Bukhari translation on sunnah.com renders it as believing in their meanings and acting accordingly. The point for anyone reciting is that memorising a list isn't obviously what the hadith asks for, and the list you'd be memorising is the weak part anyway.

So should you recite them?

Mufti Salim Qasmi, who reviews the Islamic content here: reciting the names is good and there's no harm in it. What you should not do is believe that these exact 99, in this exact order, are fixed by the Prophet, because that isn't established.

The names in the list are nearly all found in the Qur'an and in other authentic narrations. The weakness is in the enumeration, not in the names themselves. Calling on Allah by any name established in the Qur'an is beyond dispute.

What's worth dropping is the specific claim you'll see on posters and apps: that memorising this list, in this order, is the thing that earns Paradise in the hadith. The hadith doesn't say which names, and the two lists that try to supply them disagree with each other.

How far the checking went

The 4 hadith were read on sunnah.com on 10 August 2026: Bukhari 6410, Muslim 2677, Tirmidhi 3507 and Ibn Majah 3861. Gradings are the Darussalam gradings as printed there. Tirmidhi's closing comment was read in the Arabic, not from a translation.

The comparison between the two lists was done by matching the Arabic with vowel marks stripped, which is why the figures above say 'around'. Names that differ only in grammatical form were merged by hand, and that judgement is ours rather than a source's. No printed edition was consulted for any of it.

Transliterations and meanings for each name aren't here yet. Adding 99 of them from memory is exactly the thing this site doesn't do, so they'll come when each one has a source attached.