The mursal hadīth is the hadīth from which the Sahābi has been omitted. Such as when the Tabi’says that the Messenger of Allah  said or did such and such a thing, or such and such thing was done in his presence. An example would be the hadīth of a tabi'i who has met a number of companions and has sat down to learn from them like 'Ubayd Allah ibn 'Iddi ibn al-khayyar, Sa'id ibn al-Musayyab and their likes when they say (directly) that:

It is well known that all of the Tabi’een are treated equally i.e, what the tabi'i narrated about the Prophet  without mentioning the Sahabah. There is no difference between the senior or junior tabi'i because it is well known that they are treated equally. The muhaddithin, scholars of usul (usuliyyin) and the imams have differed over the use of the mursal hadīth as proof. There were those who did not use it and considered it to be rejected like the munqati' hadīth and there were those who did accept its use. Those who do not accept it reject it for a reason, which is that a transmitter who is not known has been omitted from the isnād who might not be trustworthy. The consideration in narration is reliability and certainty, an unknown transmitter is not a proof. This is the reason for rejecting the mursal ahadīth. The reason is correct and the rejection of a hadīth according to it is correct but it does not apply to the mursal hadīth because the transmitter who has been omitted is a Sahābi. Even though he is not known in terms of his identity but he is known as a Sahābi. And the Sahabah are all trustworthy ('udul). They cannot be unreliable. Rather, they are definitely trustworthy. Thus, the reason by which they reject the hadīth does not apply to the mursal and nor is there any other reason to reject it and since he fulfils the conditions of the matn, sanad and transmittor, no harm is done by omitting the Sahābi as long as it is known that he is a Sahābi and so he is trustworthy. This indicates that the mursal hadīth is a proof and should be educed as an evidence. It might be said that the reason is that there is a possibility that a tabi'i narrated from a tabi'i like himself who narrated from the Sahabah. The ommision of a Sahābi does not mean the ommision of only one narrator. But the break in the chain means that it is possible that two narrators have been omitted, one of them satisfies the condition of integrity, which is the Sahābi and the case of the other narrator is dubious, who is a tabi'i. There is a possibility in the hadīth of a jarh (invalidation) or lack of accuracy (dabt) and therefore it is rejected. Such a thing might be said. The response is that the definition of the mursal hadīth is that:

The narration of a Tabi’ from a Tabi’ who is not known does not come under this definition. Even if we accept this illustration i.e the possibility of the Tabi'i's ommision without mentioning the Sahābi, the possibility of his ommision is by way of suspicion. Rather it is a suspicion which does not reach the level of possibility. Because he suspects the tabi'i of narrating from another tabi'i who he did not mention and nor did he mention the Sahābi i.e, he assumes that a Tabi'i has been ommited. There is no evidence for this hypothetical assumption. It is merely a suspicion. A suspicion has no value and the hukm (value) of hadīth is not based on it. It should not be said that an unknown narrator (majhul) has transmitted it since the narration is not predicated on anything such that it can be said that the narrator is a majhul (unknown). Therefore, the mursal hadīth is not considered to be from the rejected ahadith, rather it is accepted and used as proof.