This document discusses regulatory validation of internal ratings-based (IRB) models used to estimate credit risk factors. It outlines some of the challenges in validating these models statistically due to limitations in historical default data and the unobservable nature of certain credit risk factors. The regulatory validation process therefore includes both qualitative and quantitative assessments. The principles of regulatory validation are described, including that it assesses predictive ability, involves independent review, and is an iterative process without a single validation method. The components of the validation process seek to demonstrate IRB systems can deliver sufficient accuracy in credit risk estimates through board oversight, use of risk rating systems in credit processes, backtesting, and stress testing.