This presentation offers a case study of the effects of legislation to protect the privacy of patient medical records on the ability to conduct public health research in the area of pharmacoepidemiology. Given the current lack of consensus regarding data privacy legislation across different states as well as intense interest at the federal level, the empirical findings of this study are of great relevance. The effect of study-specific informed consent requirements, as interpreted by one State, was dramatic and has policy implications. A post-marketing drug safety study of the relationship between tramadol and seizures in affiliated health plans required medical records for verification of case status (both cases (N=45) and matched non-cases (N=95) based on claims data). After two mailings of the consent form, 40% of the 140 members responded, with less than half of the respondents (just 16% of the total study population) agreeing to participate. After non-respondents were contacted by phone by health plan personnel, only 3 additional members agreed to participate, with a final participation rate of 19%. Those who experienced a seizure were slightly more likely to participate than non-cases. This case study illustrates the effect of study-specific consent on the ability to conduct pharmacoepidemiology research in a health plan setting and raises the policy question of how does society weigh the risk to privacy against the benefit of health research.