This paper studies the event-triggered consensus problem of second-order uncertain nonlinear multi-agent systems (MASs). Based on the local sampled measurement information, we propose an adaptive event-triggered consensus algorithm. The adaptive algorithm estimates not only the uncertain parameters of agent dynamics but also the global topology information. Hence, our consensus algorithm does not rely on global topology information, that is, the proposed consensus algorithm is full distributed. Moreover, we prove that Zeno behavior is ruled out. Finally, a simulation is given to verify the effectiveness of the proposed algorithm.