<p>This paper addresses the synchronization issue of chaotic neural networks under actuator saturation by designing a sampled-data controller. First, we propose a novel Lyapunov-Krasovskii functional consisting of looped-functionals with single and double integral terms for the error state and its derivative. Zero equations are employed so as to relax the positiveness conditions of the free matrices involved in the integral terms. Second, less conservative criteria is derived for the synchronization of chaotic neural networks with actuator saturation. Finally, we design a sampled-data controller for effective synchronization of the drive and response systems by solving linear matrix inequalities. The superiority of the proposed method are verified through a representative numerical example, showcasing its advantages over previous methods.</p>