The article introduces a simple method to estimate models with censored endogenous regressors and sample selection bias. By using generalized residuals, the approach corrects for inconsistencies caused by the endogeneity of the censored regressors. The study also presents two new tests of weak exogeneity. An example analyzing the trade-off between fringe benefits and wages shows that the estimation method is effective. A simulation indicates that the proposed tests have strong power properties.