Impulse effects on system stability are complex; namely, impulse may stabilize a system, or drive a stable system to oscillation or even chaos. We investigated the stability of impulsive cellular neural networks with time delays. By using Lyapunov functions, several conditions guaranteeing the global exponential stability of a system are established. It is shown that even if the impulsefree system is unstable, the impulsive system can be globally asymptotically stable if the impulses satisfy the conditions. A numerical example is given to illustrate the correctness of the theoretical analysis. 