In this article, we delineate the analysis of structural equation models with censored data or truncated data with the number of unmeasured subjects under the framework of the generalized censoring mechanism (GCM), one of the most popular non-ignorable missing mechanisms. An EM algorithm for maximum-likelihood estimation in structural equation models with censored data will be presented. By reformulating the censored-data problem into non-ignorable missing-data problem, the maximization procedure is shown to be simplified. A simulated data set will be used as an illustrative example. (C) 1998 Elsevier Science B.V. All rights reserved.