Traditional decision research shows that when people are told the probability of a rare event (e.g. the chance of side effects), they generally treat this event as if it were more likely than its objective probability (overweighting). In contrast, recent studies indicate that when outcomes are experienced firsthand, people tend to underweight the probability of rare events. In this paper, we suggest that the distinction between described and experienced information can play a significant role in shared decision making, and can provide a plausible explanation for some discrepancies between the perspectives of doctors and patients. We highlight some of the advantages and disadvantages of experiential and description-based information, and how knowledge of these might be used to improve risk communication.