Preventive medications such as statins are recommended to an increasingly large number of people. To those who make these recommendations, prevention is synonymous with risk reduction; the clinical task of helping people decide about preventive medication is therefore widely framed as one of risk communication. In this article, I explore the role of risk and uncertainty in accounts of medication decisions, drawing on qualitative data from interviews carried out between 2011 and 2013 in the east of England with people who had been offered statins. I found that very few participants mentioned risk or likelihood, or described weighing benefits against harms, the process central to the risk communication project. Instead, those who had decided to take statins described their certainty that statins were needed to treat current problems. This certainty was informed by knowledge about the present or the past; information about possible future harms was presented solely as contributing to concern about current problems. In contrast, those who had decided not to take statins explained their decisions in terms of the inherent uncertainty of information about the future, presenting this uncertainty as a reason to decline medication. This asymmetry between explanations for accepting and for declining statins is rooted in differences between the ways past, present and future information are handled. These findings challenge the assumption that decisions about statins are construed as decisions about risk by those offered them, and raise questions about the usefulness of using risk and uncertainty as key concepts for theoretical accounts of what is going on when people consider taking preventive medication.