Polygraph lie detectors are accurate
Polygraph machines do not reliably detect lies. They measure physiological arousal, heart rate, skin conductance, and breathing, which correlates imperfectly with anxiety in general, not specifically with deception, and can be manipulated with countermeasures; major U.S. federal scientific reviews have found their accuracy insufficient for high-stakes decisions.
What we know
Polygraph testing measures physiological indicators, heart rate, blood pressure, skin conductance (sweating), and respiration rate, and interprets patterns in these measurements around specific questions to infer whether a subject is being deceptive. The underlying assumption is that lying produces a distinct, detectable physiological signature. Decades of research have challenged this assumption directly: the measured indicators reflect general anxiety, stress, or arousal, which can be triggered by many things unrelated to deception, including fear of being wrongly accused, general nervousness in a high-stakes testing situation, or unrelated anxiety, while a practiced liar, or someone using countermeasures, can suppress or mask the physiological response the test is looking for.
The most authoritative scientific review of polygraph accuracy came from the National Academy of Sciences in a 2003 congressionally mandated report, "The Polygraph and Lie Detection." After reviewing the accumulated research literature, the committee concluded that polygraph testing, in employment screening and specific-incident investigation contexts, achieved accuracy well above chance but far below the level needed for reliable, high-stakes decision-making, and that the scientific basis for its use in broad security screening particularly (as opposed to specific, incident-focused investigations) was weak. The report specifically noted that polygraph accuracy studies conducted in real-world, high-stakes conditions are difficult to conduct rigorously (there is rarely unambiguous, independently verified ground truth about who was actually lying), which itself limits how confidently accuracy claims can be made in either direction.
Countermeasure research has repeatedly demonstrated that examinees can be trained to defeat polygraph tests, using techniques like controlled breathing, muscle tension, or mental arithmetic during control questions to artificially normalize their physiological baseline, reducing the contrast the test relies on to flag deceptive answers. Because of these limitations, polygraph results are generally inadmissible as evidence in most U.S. federal courts and many state courts, and the Employee Polygraph Protection Act of 1988 banned most private employers from using polygraph testing on employees or applicants, reflecting a legal and scientific judgment that the technology is not reliable enough for consequential decisions, even though certain government agencies, including some federal law enforcement and national security roles, continue using polygraphs for screening purposes under different legal standards and with acknowledged limitations.
The persistence of public belief in polygraph accuracy stems partly from decades of dramatic media and television portrayals, and partly because the test can still function as a genuinely useful interrogation tool, not because it reliably reads truthfulness, but because the belief that it works can pressure some subjects into confessions or inconsistent statements during the broader interview process, an effect psychologists distinguish from the polygraph actually detecting lies through physiology. The Central Intelligence Agency and Federal Bureau of Investigation continue using polygraph testing for internal security screening purposes despite these documented limitations, a practice critics argue is maintained partly for its psychological deterrent effect on job applicants and personnel rather than because the underlying physiological measurement itself has been shown to reliably detect deception.
Common claims
- Polygraphs reliably detect lying with 90%+ accuracy.False. The NAS found accuracy well below perfection, especially for screening.
- Countermeasures cannot fool a polygraph.False. Physical and cognitive countermeasures can defeat the test under lab conditions.
- Passing a polygraph proves innocence.False. Innocent people regularly fail, and guilty people regularly pass.