Proposed in this paper is a novel channel-hopping algorithm for rendezvous on the basis of asymmetric model in cognitive networks (named AMCH)． According to the available channel set sensed by each secondary user，the lengths of channel hopping sequences for transmitters and receivers are selected in two disjoint sets of prime numbers，and two channel hopping sequences are respectively constructed． Without any other auxiliary infor-mation (such as the synchronization information and the global channel label information，etc． )，any two secondary users within the communication scope can rendezvous in a short time and establish a link in their common channels．In comparison with the existing algorithms，AMCH algorithm diminishes the redundancies of channel-hopping se-quences and improves the efficiency of rendezvous． Simulated results show that AMCH algorithm outperforms the existing algorithms in terms of time-to-rendezvous upper boundary and rendezvous performance．