This paper investigates the delay-probability-distribution-dependent stability problem for a class of neural networks with time-varying delays. The probabilistic delay satisfies a certain probability distribution. By introducing a stochastic variable with a Bernoulli distribution, the neural networks with random time delays is transformed into one with deterministic delays and stochastic parameters. Based on the Lyapunov-Krasovskii functional, a novel delay-probability-distribution-dependent sufficient condition in the form linear matrix inequality (LMI) such that delayed neural networks are globally asymptotically stable in the mean square. Numerical examples are given to illustrate the effectiveness of the proposed method. © 2012 Watam Press.