I discuss gravitational wave experimental signatures (bursts and stochastic background) of cosmic strings. I will show burst rates that are substantially lower (about a factor of 1000) than previous estimates suggest and explain the disagreement. Initial LIGO is unlikely to detect bursts from field theoretic cosmic strings, though it may detect cosmic superstring bursts. I also compare the stochastic background produced by a network of strings with a wide range of experiments and indirect bounds. If the latest cosmic string simulation results are correct then a large area of superstring parameter space is ruled out by pulsar timing observations.