A lie detection test, most commonly a polygraph exam, is a procedure that monitors your body’s involuntary physical responses while you answer a series of questions. The idea is straightforward: when someone lies, the effort of concealing the truth triggers measurable changes in breathing, heart rate, blood pressure, and sweat gland activity. An examiner then interprets those changes to judge whether your answers were truthful or deceptive. The polygraph remains the most widely used form of lie detection, though its accuracy and scientific standing are far from settled.
What a Polygraph Actually Measures
A polygraph records data from five sensors attached to your body. Two elastic bands around your chest and abdomen track your breathing rate and depth. A standard blood pressure cuff on your upper arm monitors cardiovascular activity, including heart rate and blood pressure changes. Two small metal plates placed on your fingertips measure electrodermal activity, which is the slight increase in electrical conductivity that occurs when your sweat glands activate. A fifth sensor, typically a pad beneath your seat or feet, is designed to detect deliberate muscular movements that might be used to manipulate the results.
None of these sensors measure lying directly. They measure arousal, the body’s general stress response. The core assumption is that a deceptive answer produces a stronger stress response than a truthful one. This distinction matters because many things besides lying can trigger the same physical reactions: anxiety, anger, confusion, or simply being nervous about taking the test itself.
The Three Phases of the Exam
A full polygraph examination typically runs two to three hours and follows a structured three-phase process.
The pre-test interview is the longest phase, often lasting 45 to 90 minutes on its own. The examiner explains how the instrument works, discusses the issue under investigation, and goes over every question that will be asked during the actual test. Nothing is sprung on you. You’ll also complete paperwork, be informed of your rights, and be assessed for any physical or psychological conditions that might affect the results. The entire session is usually audio and video recorded.
During the in-test phase, you’re connected to the polygraph and the blood pressure cuff is inflated to a moderate pressure. The examiner runs a brief calibration test, then asks the predetermined questions while the instrument records your physiological responses. This is typically done in multiple rounds, called “charts,” to gather enough data for analysis.
In the post-test phase, the examiner reviews and scores the charts. Scoring can be done by hand using a numerical system, where the examiner assigns positive or negative values to each response, or by computer algorithm. The most current automated system, known as the Objective Scoring System, has been shown in research to outperform human scoring in detecting guilty participants. Some examiners use both methods as a cross-check.
How Accurate Is It?
This is where things get complicated. The most authoritative review of polygraph science came from the National Research Council in 2003, which examined decades of research and found a median accuracy index of 0.86 on a scale where 1.0 would be perfect and 0.50 would be random chance. In practical terms, this means the polygraph performs well above guessing but falls short of the reliability expected from forensic evidence.
The same National Research Council report identified “major deficiencies” in polygraph technology. The fundamental problem is that there is no unique physiological signature of lying. The test detects arousal, and arousal has many causes. An innocent person who is anxious about being accused can produce the same readings as someone who is actually being deceptive. This means the polygraph generates both false positives (truthful people flagged as liars) and false negatives (liars who pass the test).
Countermeasures are another concern. Research published in the Journal of Applied Psychology found that people trained in simple physical techniques, like pressing their toes to the floor or biting their tongue during control questions, or mental techniques like counting backward by seven, could defeat the polygraph about 50% of the time. These countermeasures were difficult for examiners to detect either by observation or through the instrument’s data.
Legal Status in the United States
Polygraph results occupy an unusual legal position. Neither the U.S. Code nor the Federal Rules of Evidence specifically address their admissibility, and military courts have barred them entirely since 1991. For decades, most federal courts excluded polygraph evidence under the “general acceptance” standard from a 1923 case, reasoning that the scientific community had not broadly accepted the technique as reliable.
A 1993 Supreme Court ruling loosened that standard, requiring judges to evaluate scientific evidence based on whether it has been tested, peer-reviewed, has a known error rate, and has gained acceptance in the relevant scientific community. Since then, a small number of federal circuit courts have opened the door to polygraph evidence in narrow circumstances, such as when both parties agree to the test beforehand, or when results are used to support or challenge a witness’s credibility. In most courts, however, polygraph results remain inadmissible.
In the workplace, the Employee Polygraph Protection Act of 1988 prohibits most private employers from requiring or even requesting that employees or job applicants take a lie detector test. There are specific exceptions: federal, state, and local government agencies are not covered by the ban. Private-sector exemptions exist for security firms (armored car companies, alarm companies, security guard services), pharmaceutical companies where employees handle controlled substances, and situations where an employer is investigating a specific workplace theft or loss and the employee had access to the property in question.
Alternative Lie Detection Technologies
Voice stress analysis emerged as a potential alternative to the polygraph, designed to detect microtremors in a person’s voice caused by the effort of concealing the truth. Unlike the polygraph, voice stress analysis software can be applied to phone calls, recordings, and remote conversations without requiring a person to sit in a room connected to sensors. In practice, however, the technology has performed poorly. A field study funded by the National Institute of Justice found that voice stress analysis programs correctly identified only about 15% of people who lied about recent drug use (verified against urine tests). One of the two programs tested caught just 8% of deceptive responses. While they were better at correctly identifying truthful people (about 91.5% accuracy on that measure), a test that misses 85% of actual lies has limited practical value.
Brain imaging using functional MRI represents a more technologically sophisticated approach. Researchers have identified patterns of activity in decision-making and emotional regions of the brain that differ between truthful and deceptive responses. However, this technology remains in the experimental stage. There have been no prospective clinical trials meeting the standards required for medical device approval, and the few studies conducted have produced inconsistent activation patterns across subjects. Some participants showed the expected frontal brain activity during deception, while others showed emotional responses in completely different brain regions, suggesting that people may lie in neurologically different ways.
What to Expect if You Take One
If you’re asked to take a polygraph for employment screening or as part of an investigation, the process is voluntary in most civilian contexts. You cannot be forced to take one, and in many jurisdictions, refusing a polygraph cannot legally be held against you in employment decisions covered by the EPPA.
The sensors are painless. The blood pressure cuff inflates to a moderate level (about 60 mmHg, roughly half of what you’d feel during a standard blood pressure check at a doctor’s office), and the finger plates simply rest against your skin. You’ll sit still in a chair and answer “yes” or “no” to questions you’ve already reviewed. The examiner will run through the question set multiple times to collect enough data.
After the test, the examiner scores the charts and renders an opinion: no deception indicated, deception indicated, or inconclusive. That opinion is exactly that, an interpretation of physiological data filtered through a scoring method that, even under favorable conditions, has a known margin of error. The result is not a definitive statement about whether you told the truth.