A series of experiments illustrates a learning process that enhances brand equity at the expense of quality-determining attributes. When the relationship between brand name and product quality is learned prior to the relationship between product attributes and quality, inhibition of the latter may occur. The phenomenon is shown to be robust, but its influence appears sensitive to contextual variations in the learning environment. Tests of process are inconsistent with attentional explanations and popular models of causal reasoning, but they are supportive of associative learning models that portray learners as inherently forward looking. P urchase decisions are based on predictions of product performance. Consumers base their predictions in part on product cues and are accurate to the extent that they have properly learned the relationship between the cues and performance. Consumer research has devoted little attention to this learning process despite its fundamental importance (Hutchinson and Alba 1991; Meyer 1987). In the present research we examine consumer learning of product cues as predictors of product quality with particular emphasis on the distinction between brand and attribute cues. To illustrate, consider the cases in which consumers rely strictly on either brand or attribute cues to predict quality. If consumers learn the relationship between product attributes and quality, they will differentiate among brands that possess different attributes and treat as commodities those brands that share the same attributes. Once the predictive rule is learned, it may be applied to any new brand that possesses the attributes. In contrast, consumers who rely strictly on brand cues will ignore the underlying attributes and may incorrectly differentiate physically identical brands. The latter case is important because it can be costly and is not uncommon (such as when consumers pay high premiums for branded drugs that are chemically identical to their generic counterparts). An appealing explanation of this phenomenon is that consumers are unaware of the attributes of these brands. Indeed, firms attempt to foster such ignorance by making attribute information difficult to find or process (Bergen, Dutta, and Shugan 1996; Hoch and Deighton 1989). The present research investigates whether consumers will routinely learn the determinants of product quality when attribute cues are *Stijn M. J. van Osselaer is assistant professor of marketing, University of Chicago. Joseph W. Alba is distinguished professor of marketing, University of Florida. This work was supported by the Beatrice Foods Co. Faculty Research Fund at the University of Chicago Graduate School of Business. The authors thank Alan Cooke, Joffre Swait, Chris Janiszewski, and Bart Weitz for their helpful comments. freely available and processing is unconstrained. We suggest that learning can be suppressed even under these relatively favorable conditions due to the learning phenomenon known as blocking.