The effective implementation of Quality by Design (QbD) for wet granulation requires quantitative models for design and scaling of granulators. This paper reviews that current state of the art of granulation modeling, especially with respect to QbD in the pharmaceutical industry. Our thesis is that the state of the art use of high shear wet granulation models for QbD should be at the intermediate level between the extremes of data-driven empirical models and first principles models, i.e., regime-based models. These models have a sound mechanistic basis and can be used for both formulation design and process scaling. Although not fully predictive, regime map approaches are a significant advancement over a typical design of experiments approach, where little is known about the process. In addition, regime maps are particularly useful for process scale-up/down. This review focuses on the status of wet granulation regime maps and critically assesses their applicability to real, rather than model formulations. The approach is illustrated by two case studies. The first case study evaluates the effect of changing the properties of one excipient, lactose, on formulation behavior. Changing lactose size distribution leads to substantial changes in both the formulation drop penetration time and dynamic yield stress, which can lead to changes in operating regime. A size measurement technique that accurately measures the fine end of the distribution is necessary to predict the effect of changing the lactose size distribution. The second case study uses the regime map approach to propose a design space and scaling rules for a high drug load formulation, and successfully validate the scaling rules for scaling from a 10 l to a 75 l granulation. This case study emphasizes the importance of spray distribution to avoid operation in the mechanical dispersion regime and subsequent lump formation as a granulation is scaled up. (C) 2012 Elsevier B.V. All rights reserved.