The sample included 60 white male students who ranged in age from 18 to 22 years and who were treated as guilty or innocent of a mock theft. Each subject was polygraphed using the Modified General Question Test which contains two control questions, five relevant questions, and four irrelevant questions. With the exception of control questions, the question list was identical for all subjects. Numerical scorings of polygraph data showed that innocent persons were more responsive to control than to relevant test questions. In addition, volunteers' subjective ratings of concern for control and relevant questions were strongly consistent with field observations and prior similar assessments. The objective measurement of skin resistance response, generally the most effective physiological measure in laboratory-based research, also showed support for the expected relationship between subjects' guilt and differential responses to control and relevant test questions. The findings show that control question testing, relative to testing without control questions, has empirically demonstrated advantages. Control question testing enhances the ability to use physiological data to discriminate more effectively between truthful and deceptive subjects. Moreover, the use of control questions reduces the probability of false positive errors. 26 references, 6 tables.