A nominative signature scheme allows a nominator (or signer) and a nominee (or veri¯er) to jointly generate and publish a signature in such a way that only the nominee can verify the signature and if nec- essary, only the nominee can prove to a third party that the signature is valid. In a recent work, Huang and Wang proposed a new nominative signature scheme which, in addition to the above properties, only allows the nominee to convert a nominative signature to a publicly veri¯able one. In ACISP 2005, Susilo and Mu presented several algorithms and claimed that these algorithms can be used by the nominator to verify the validity of a published nominative signature, show to a third party that the signature is valid, and also convert the signature to a publicly veri¯able one, all without any help from the nominee. In this paper, we point out that Susilo and Mu's attacks are actually incomplete and in- accurate. In particular, we show that there exists no e±cient algorithm for a nominator to check the validity of a signature if this signature is generated by the nominator and the nominee honestly and the Decisional Di±e-Hellman Problem is hard. On the other hand, we point out that the Huang-Wang scheme is indeed insecure, since there is an attack that allows the nominator to generate valid nominative signatures alone and prove the validity of such signatures to a third party.