The Australian Building Greenhouse Rating (ABGR) scheme is a performance-based rating for office buildings. While it was originally designed for existing buildings, its success in that sector has led to its rapid adoption as a performance standard for new buildings. As a result of this application, there has been an upsurge in the use of simulation as a predictor of absolute performance. The problems inherent in such absolute predictions led to the creation of the ABGR Validation Protocol for Computer Simulations. This paper describes the background and content of the Protocol, and highlights issues and lessons learnt from its development and application.