We provide a review and a theoretically unifying model of framing effects. Beginning with a review of findings that prospect theory, fuzzy‐trace theory, and traditional dual process models do and do not explain, we develop a model of framing based partly on Bless and Schwarz's (2010) inclusion/exclusion model of assimilation and contrast. Our model is the first of its kind to be applied to all three valence framing manipulations and presents common mechanisms explaining findings in the framing literature that were previously difficult to conceptualize within a single theory (e.g., matching effects and cognitive effort). The model also accounts for why all three framing manipulations are influenced by differences between holistic/contextual versus analytic processing as well as the presence, absence, and direction of effects produced by different versions of framing manipulations. It also predicts contrast effects (Implications and Future Directions section) and applies to manipulations involving a frame and judgmental target. Copyright © 2017 John Wiley & Sons, Ltd.