Model validation of power plant equipment is expected to be mandated imminently by the North American Electric Reliability Corporation. These standards will result in a great need for performing model validation for a large fleet of existing and future power plants. The historic approach to model validation of synchronous generators has typically been a battery of offline tests. In addition, the process of validation depends on a trial and error series of simulations, comparing the results of such simulations with measured test data, to gradually tend toward a validated model and its parameters. Such a method presents significant challenges for application on a large fleet of units. This paper documents a practical approach to model validation. The method is based on collecting online data from the generating unit, typically near or at base-load. Data are collected either from digital-fault recorders capturing system disturbances, or from digital excitation system for an imposed small voltage-reference step. The data are then used in a software tool to perform model validation using an automated optimization algorithm-the entire process may be said to be semiautomated. The benefits are explained in the paper through actual examples of the application.