This study empirically investigated bootstrap bias estimation in the area of structural equation modeling (SEM). Three correctly specified SEM models were used under four different sample size conditions. Monte Carlo experiments were carried out to generate the criteria against which bootstrap bias estimation should be judged. For SEM fit indices, bias estimates from the bootstrap and Monte Carlo experiments were quite comparable in most cases. It is noted that bias was constrained in one direction in the Monte Carlo experiments because of the perfect fit of the true SEM models. For the SEM loadings and coefficients, the difference between bootstrap and Monte Carlo bias estimations was very small, and the distributions of the bias estimators from the two experiments were quite similar. For the SEM variances/covariances, the comparison of the bias estimator distributions from the two experiments indicated that bootstrap bias estimation could be considered adequate. Because the study involved three SEM models which served as an internal replication mechanism, the likelihood of chance discovery for the findings was small, and the findings should have reasonable generalizability. Future studies may extend the current findings by examining misspecified SEM models. Data nonnormality may be another dimension to be considered in future investigations. (Contains 6 figures, 4 tables, and 40 references.) (Author)