OBJECTIVES: To compare side effect data reported from members of a patient-registry to information contained in the product labeling of four different cholesterol medications each representing a different therapeutic class. METHODS: A random sample of U.S. MediGuard.org members who reported taking niacin (Niaspan), fenofibrate (Tricor), simvastatin (Zocor), or colesevelam (Welchol), were invited to complete a validated online treatment satisfaction survey that includes questions related to side effects.  MediGuard.org is a free medication monitoring service that provides information to over 2.5 million members in the US, UK, France, Germany, Spain, and Australia.  A comparison list of adverse effects and frequencies reported during clinical trials was extracted from the branded package insert for each medication.  RESULTS: Feedback was obtained from 56 colesvelam, 108 niacin, 216 fenofibrate, and 660 simvastatin patients.  Niacin patients had the highest prevalence of side effects (62%), primarily flushing (55%) and pruritis (14%).  23% of colesevelam members reported side effects: the most common were constipation (14%) and bloating/gas (2%).  For patients treated with fenofibrate, 16% reported side effects with myalgia (4.2%) and arthralgia (1.4%) being the most common. Similar to fenofibrate, 16% of simvastatin patients reported side effects and again, myalgia (6.2%) and fatigue (2.1%) had the most mentions. Finally, 2% of niacin, 22% of fenofibrate, 29% of simvastatin, and 36% of colesevelam patient reports included side effects not included in the product labeling. CONCLUSIONS: On-line patient communities are an emerging resource for confirming adverse events reported during clinical trials and for capturing previously undocumented signals. In tandem to the current post-marketing spontaneous adverse event reporting system, longitudinal patient registries can provide insight not only on the number of adverse events, but also a prevalence rate of those who experience versus those who do not experience side effects.