Occasions · 7 min read
The Laylatul Qadr dua, and the word most posters add
The dua taught for Laylatul Qadr is the most shared dua of Ramadan, and it's the one most often printed with an extra word in it. This page gives the established wording with its Arabic, its source and the grading that can be named, then shows where "kareem" came from and what the scholars who examined it said. The dua is a live preset in the counter here, and the last section says what that preset carries today and what it's still missing.
What is the dua for Laylatul Qadr?
The Laylatul Qadr dua is: Allahumma innaka 'Afuwwun tuhibbul-'afwa fa'fu 'anni. It means "O Allah, You are Pardoning, You love pardon, so pardon me." The Prophet taught it to A'ishah when she asked what to say if she found the night. It's recorded in Jami' at-Tirmidhi 3513.
On the grading, here's what can be pointed at and what can't. Tirmidhi 3513 sits outside the two Sahihs, so "authentic hadith" with no qualifier won't do. The grading with a name attached is sahih per Ali Za'i, printed on hadithunlocked. That's the one printed here.
The phrase "hasan sahih" gets quoted alongside this hadith constantly, and the checking behind this page couldn't tie it to a named grader, so it isn't printed here as anyone's verdict. The same happened with Tirmidhi 3575 on the morning and evening adhkar: a "hasan sahih by Tirmidhi" attribution went unverified and was dropped.
One gap worth naming. This site's rule is to print the sunnah.com grade next to a scholar's whenever they differ, and sunnah.com's grade for 3513 isn't on record here. Open the link and read it. That care matters, because what follows is a dispute about wording.
- Arabic: اللَّهُمَّ إِنَّكَ عَفُوٌّ تُحِبُّ الْعَفْوَ فَاعْفُ عَنِّي
- Transliteration: Allahumma innaka 'Afuwwun tuhibbul-'afwa fa'fu 'anni
- English: O Allah, You are Pardoning, You love pardon, so pardon me
- Source: Jami' at-Tirmidhi 3513, at sunnah.com/tirmidhi:3513
- Grading: sahih per Ali Za'i, printed on hadithunlocked
- Count: no number narrated, repeated freely through the last ten nights
Why does the word kareem appear on most versions of this dua?
The word kareem isn't in the established wording. Shaykh Ibn 'Uthaymin's position, quoted by Shaykh al-Munajjid's IslamQA in answer 260806, is that it entered through a scribe or a modern typesetter. Dozens of hadith books carry the dua without it, and al-Albani corrected his own earlier printing.
The extra word is كَرِيمٌ, kareem, meaning generous. In the version on the posters it sits alongside 'Afuwwun, so the opening reads "You are Pardoning, Generous". The word itself carries no difficulty. The question is only whether this narration transmits it.
Al-Albani's correction is the detail worth sitting with. The word appeared in an earlier printing of his own Silsilat al-Ahadith as-Sahihah and he took it out. A scholar correcting his own book against the manuscripts is the strongest form this kind of evidence takes.
Now the thing you'll find if you go and check, which you should. sunnah.com's page for Tirmidhi 3513 prints "[Karimun]" in square brackets. The brackets mark a variant: the word is present in some manuscript editions, and the site is flagging it rather than transmitting it as settled text. So the reference this page sends you to does show the disputed word. Saying that here, instead of letting you find it and conclude the correction was careless, is the reason this page exists.
Is it wrong to say the dua with kareem in it?
Saying the dua with kareem is not something this page rules on, and none of the sources cited here rules on it either. Ibn 'Uthaymin's objection is about what the narration transmits, not about the meaning of the word. The shorter wording is what's established.
State it in the direction the sources support. What's established is the shorter dua: that's what the collections transmit, and that's where Ibn 'Uthaymin and al-Albani landed. On the other side there's a manuscript variant that sunnah.com brackets, with no named scholar arguing the longer version back into the hadith. Those carry different weight and shouldn't be printed as equals.
Everything cited here is about transmission. If you've said it with kareem for years, what these sources address is which words the narration carries, and nothing beyond that. Correcting a wording says nothing about whether the practice was right. The night, the dua and the asking are all established. It's one word.
When is Laylatul Qadr, and when is this dua said?
Laylatul Qadr falls in the last ten nights of Ramadan and its exact night isn't fixed in advance. That's the setting of the hadith itself: A'ishah asks what to say if she finds the night, not how to identify it. The dua is repeated freely across those nights.
The frame of the narration is the useful part. A'ishah's question is conditional, and the answer she's given is a dua rather than a sign to watch for. Read that way, the wording is built for someone searching the last ten nights without certainty, which is everybody.
This page prints no hadith number for the narrations about the odd nights, because none was checked for it. Every number on this page was checked against the collection before it went up, and those weren't part of that check. Look them up in a collection directly rather than taking a number off a page that didn't verify it, this one included.
As for what the night is for, the dua answers that better than a commentary would. It asks for no provision, no health, no list. It addresses Allah as al-'Afuww, the Pardoning, says that He loves to pardon, and then asks for pardon.
How many times should the Laylatul Qadr dua be said?
No count is narrated for the Laylatul Qadr dua. The hadith gives the wording and says nothing about how many times. Any figure printed on a graphic is somebody's choice, not an instruction from the narration. Repeat it as much as you want through the last ten nights.
The silence is in the text, and it reads clearly against the dhikr where a number is narrated: 33 each after fard salah in Sahih Muslim 597a, 100 for istighfar in Sahih Muslim 2702, 3 for Juwayriyah's phrase in Sahih Muslim 2726a. There the count comes from the narration. Here it doesn't, so nothing on this site prints one.
Scholars differ on choosing your own number. Shaykh al-Munajjid's IslamQA holds that restricting an adhkar to a specific number the texts don't mention falls under bid'ah, while exceeding a narrated number, or doing an unrestricted dhikr a lot, is fine. SeekersGuidance is more permissive about a personally chosen daily wird. Both positions are on the record and this page doesn't rule between them.
What this counter ships, and what it doesn't yet
The Laylatul Qadr dua ships today. It's one of 21 presets in the counter, carrying the wording above without kareem, with no target set, and with Jami' at-Tirmidhi 3513 linked from the entry to sunnah.com. What isn't built yet is the wider Ramadan and Laylatul Qadr program pack.
The no-target part is checkable in the app. 8 of the 21 presets carry no count at all: the field is empty, the counter prints "no set count" where a target usually sits, and you count until you stop. The two heavy words, from Sahih al-Bukhari 6682, sits in that same group with the note "The hadith gives no count. 100 is common practice, not a narrated number." The Laylatul Qadr dua is there too, rather than having a number invented for it.
The entry's own note already carries the correction. It names IslamQA 260806 and Ibn 'Uthaymin, records that dozens of hadith books transmit the dua without the word, and says al-Albani corrected his own earlier print. Two things are missing from it. It doesn't show the bracket disclosure from the second section, and it prints no grading, which this site's rule requires for a narration outside Bukhari and Muslim. Both are open and both are listed below.
The program pack is the larger gap. One guided program ships today, the post-salah sequence. The last-ten-nights and tahajjud sets aren't built yet.
If any of this turns out to be wrong, it goes on the corrections page with the date, what was published, and what it says now. That log opened on 5 August 2026 and it's empty, which says the site is new rather than that it's free of errors.
- Shipped: the wording without kareem, copied from the collection rather than typed from memory
- Shipped: no target, because no count is narrated
- Shipped: Jami' at-Tirmidhi 3513 on the entry, linked to sunnah.com
- Shipped: the note naming IslamQA 260806, Ibn 'Uthaymin, and al-Albani's correction of his own print
- Not yet: a line saying sunnah.com prints "[Karimun]" in brackets, and why
- Not yet: the grading, sahih per Ali Za'i, on the entry itself
- Not yet: the Ramadan and Laylatul Qadr program pack the dua belongs in