The objective of this paper to analyze dissipativity of discrete-time neural networks with time-varying delay. The main idea is to introduce the concept of extended dissipativity for discrete-time neural networks with a view to unifying several performance measures such as the H∞ performance, passivity, l2-l∞ performance and dissipativity. The reciprocally convex approach together with a Lyapunov function involving a triple-summable term is applied to develop the extended dissipativity criterion for discrete-time neural networks with time-varying delay. In addition, the new criterion also ensures the stability of the neural networks. The improved results are validated through a numerical example in comparison with the existing results.