In the presence of stochastic perturbations and time delay, the aperiodic event-triggered synchronization control for neural networks is studied in this brief. First, a suitable linear event-triggered control is designed to ensure the synchronization of the delayed stochastic neural networks in mean square. Then, by using Itô-differential formula, the synchronization criteria are derived under the event-triggered control. Moreover, a new method is proposed to prove that Zeno behavior does not occur, which ensures that the event-triggered control is feasible. Finally, an example and its simulation are provided to substantiate the effectiveness of the theory.