Prison-based research has been limited due to concern that prisoners may represent a vulnerable population secondary to possible coercion and limited capacity for voluntary informed consent. This study was designed to assess decisional capacity and susceptibility to coercion in prison research subjects. Subjects were 30 mentally ill prisoners and 30 healthy controls. The groups were compared on ability to provide informed consent to a hypothetical drug trial, susceptibility to possible coercion, neuropsychological functioning, and psychiatric symptoms. Results indicated that all controls and all but one of the prisoners demonstrated adequate capacity to consent to the hypothetical drug trial. However, when decisional capacity was measured quantitatively, prisoners performed significantly worse regarding two aspects of this ability. Regarding possible coercion, prisoners' main reasons for participating in research included avoiding boredom, meeting someone new, appearing cooperative in hopes of being treated better, and helping society. Neuropsychological functioning was strongly positively correlated with decisional capacity and negatively correlated with susceptibility to possible coercion, whereas psychiatric symptoms were only weakly correlated with these variables. In conclusion, a very high percentage of particularly vulnerable, mentally ill prisoners demonstrated adequate capacity to consent to research. Lower scores on a quantitative measure of decisional capacity suggest that extra care should be taken during the consent process when working with these subjects. The reasons prisoners gave for participating in our research indicated that the prison setting may have influenced their decision to participate, but that they were not actually coerced into doing so. Despite serious past incidents, ethicists will need to consider the possibility that prisoners have become an overprotected population.