A class of rank statistics for testing a change in distribution of randomly censored data is presented. The limit behaviour of these test statistics under null hypothesis of no change is described.