Although the need to obtain "informed" consent is institutionalised as a principle of ethical practice in research, there is persistent evidence that the meanings people attribute to research tend to be substantially at variance with what might be deemed "correct". One dominant account in the ethics literature has been to treat apparent "misunderstandings" as a technical problem, to be fixed through improving the written information given to research candidates. We aimed to explore theoretically and empirically the role of written information in "informing" participants in research. We conducted a qualitative study involving semi-structured interviews with 29 unpaid healthy volunteers who took part in a genetic epidemiology study in Leicestershire, UK. Data analysis was based on the constant comparative method. We found that people may make sense of information about research, including the content of written information, in complex and unexpected ways. Many participants were unable to identify precisely the aim of the study in which they had participated, saw their participation as deriving from a moral imperative, and had understandings of issues such as feedback of DNA results that were inconsistent with what had been explained in the written information about the study. They had high levels of confidence in the organisations conducting the research, and consequently had few concerns about their participation. These findings, which suggest that some "misunderstanding" may be a persistent and incorrigible feature of people's participation in research, raise questions about the principle of informed consent and about the role of written information. These questions need to be addressed through engagement and dialogue between the research, research participants, social science, and ethics communities.