"Count the crescent of Shaʿbān for Ramadan and do not mix it with Ramadan unless it coincides with a fast that one of you was observing. Fast upon seeing it and break the fast upon seeing it. If it is obscured from you, then the count is not obscured from you." Narrated by al-Dāraqutnī (p. 230) and al-Ḥākim (1/425), and from them by al-Bayhaqī (4/206) and al-Baghawī in "Sharḥ al-Sunnah" (2/182/1-2) through the chain of Abū Muʿāwiyah from Muḥammad ibn ʿAmr from Abū Salamah from Abū Hurayrah who said: The Messenger of Allāh (ṣallá Allāhu ʿalayhi wa-sallam) said it. Al-Ḥākim said: "Authentic according to the criteria of Muslim." Al-Dhahabī agreed with him. I say: It is only ḥasan due to the disagreement regarding Muḥammad ibn ʿAmr and because Muslim did not use him as evidence but only narrated from him as corroboration. Moreover, the complete ḥadīth is only with al-Dāraqutnī, and al-Baghawī does not have the statement: "Fast upon seeing it..." etc. Al-Ḥākim only has the first part of it. Likewise, it was narrated by al-Tirmidhī (1/133) and he criticized it by saying: "We do not know it like this except from the ḥadīth of Abū Muʿāwiyah, and the authentic version is what was narrated from Muḥammad ibn ʿAmr from Abū Salamah from Abū Hurayrah from the Prophet (ṣallá Allāhu ʿalayhi wa-sallam) who said: 'Do not precede Ramadan by a day or two.'" Similarly, it was narrated from Yaḥyá ibn Abī Kathīr from Abū Salamah from Abū Hurayrah from the Prophet (ṣallá Allāhu ʿalayhi wa-sallam) similar to the ḥadīth of Muḥammad ibn ʿAmr al-Laythī. I said: Since al-Tirmidhī only encountered the first part of the ḥadīth as we indicated, he assumed that Abū Muʿāwiyah made a mistake in it and said: "Count the crescent of Shaʿbān for Ramadan" instead of saying: "Do not precede..." etc. Therefore, he judged it as an error, but I do not see it that way because the narration of al-Dāraqutnī has combined between..