Abstract: The rise of big data and sophisticated, machine learning algorithms is increasing the prevalence of price discrimination and even personalized pricing. In traditional models, where consumers’ willingness-to-pay (WTP) is a function of preferences (and budget constraints), price discrimination is often celebrated for increasing efficiency albeit while reducing consumer surplus. This favourable view of price discrimination should be re-evaluated when WTP is a function of both preferences and misperceptions. With demand-inflating misperceptions, price discrimination is even more harmful to consumers and might reduce efficiency. These results are derived using a simple, linear demand model with different levels of price discrimination (or segmentation). In the many consumer markets where misperception is common, more careful scrutiny of price discrimination is warranted.