Two-party authenticated key agreement protocol is used to authenticate entities and establish session keys in an open network in order to provide secure communications between two parties. Several security attributes are highly desired for such protocols, such as perfect forward secrecy (the corruption of long-term keys of all the entities should not compromise any session key), PKG forward secrecy (the corruption of the PKG's master key in the ID-based system should not compromise the established session keys), and known session-key specific temporary information secrecy (The exposure of private temporary information should not compromise the secrecy of generated session keys). In 2005, Choie et al. proposed three identity-based authenticated key agreement protocols from pairings. Our analysis shows that they all didn't provide protection against known session-key specific temporary information attack and some of them are vulnerable against man-in-the-middle attack, such as the key replicating attack. We analyze some of the attacks under the BR93 security model.