Stein-rule procedure is a known technique for yielding biased but efficient estimators of parameters. This article demonstrates that it can be utilized for overcoming the inconsistency of least squares estimators in measurement error models and therefrom providing a class of consistent estimators.