A polygraph test produces a series of squiggly lines on a chart (or, more often now, on a computer screen), and each line represents a different body function recorded in real time. Reading the results means comparing how your body responded to different types of questions, looking for spikes, dips, and shifts that suggest heightened stress during specific moments. Here’s how that process works from start to finish.
What the Sensors Actually Measure
A polygraph records three core physiological signals simultaneously. Rubber tubes called pneumographs wrap around the chest and abdomen to track the rate and depth of breathing. A standard blood pressure cuff sits over the bicep to measure relative blood pressure and heart rate. Electrodes attached to two fingertips or the palm pick up electrodermal activity, which is essentially how much your sweat glands are firing at any given moment.
Each sensor feeds its data to a separate channel, producing its own line on the chart. So when you look at a polygraph readout, you’re seeing at least three parallel tracings running left to right across the page, each one telling a different part of the body’s stress story.
How Questions Are Structured
Before any sensors are attached, the examiner conducts a lengthy pre-test interview. This serves two purposes: it helps the examiner understand the facts of the case well enough to write clear, unambiguous questions, and it gives them a chance to observe the subject’s baseline behavior and quirks that might affect the results. The interview also, by design, reinforces the subject’s belief that the machine will catch a lie, which can amplify physiological reactions driven by anxiety.
During the actual test, three types of questions appear in a carefully planned sequence. Irrelevant questions (“Is today Tuesday?”) establish a calm baseline. Control questions are designed to make almost everyone uncomfortable, typically asking about minor past dishonesty (“Have you ever lied to someone who trusted you?”). Relevant questions address the specific issue under investigation (“Did you take the money from the safe?”). The entire logic of polygraph interpretation rests on comparing how the body reacts to control questions versus relevant questions.
Reading the Breathing Tracings
The respiratory channels typically appear as smooth, wave-like lines at the top of the chart. Each crest represents an inhale; each trough, an exhale. A relaxed person produces a steady, rhythmic pattern with consistent height (amplitude) and spacing.
Examiners look for several changes after a relevant question is asked. A sudden decrease in amplitude, where breaths become noticeably shallower, is one common indicator. A shift in baseline, meaning the entire wave drifts upward or downward from where it had been sitting, is another. Some people hold their breath briefly or switch to an irregular rhythm. The target amplitude for normal respiratory activity is roughly three-quarters of an inch on a traditional chart, so deviations from that range stand out. When breathing looks the same after both control and relevant questions, that channel is scored as showing no meaningful difference.
Reading the Sweat Response
The electrodermal channel usually appears lower on the chart. When there’s no significant reaction, this tracing runs nearly as a flat horizontal line. When the sweat glands activate in response to stress, the line rises in a sharp upward deflection and then gradually slopes back down.
Two things matter here: how tall the spike is (amplitude) and how long it takes to return to baseline (duration). A larger, longer-lasting response after a relevant question compared to a control question is interpreted as a sign of deception. Stronger emotional stimulation produces bigger responses, but repeated exposure to the same stimulus causes the reaction to fade over time, a phenomenon called habituation. This is one reason examiners vary question order across multiple rounds of testing.
Reading the Cardiovascular Tracing
The blood pressure and pulse channel records the rhythmic squeeze and release of the heart. Each heartbeat creates a small upward stroke (the systolic pulse, when the heart contracts) followed by a downward stroke (the diastolic phase, when it relaxes), with a small notch in between called the dicrotic notch. This gives the tracing a sawtooth appearance.
Examiners evaluate four features in this channel: changes in baseline (the overall position of the tracing shifting up or down), changes in amplitude (the height of each pulse wave growing or shrinking), changes in rate (heartbeats speeding up or slowing down), and changes in the dicrotic notch, including its position shifting or disappearing entirely. During a fight-or-flight response, heart rate and blood pressure both tend to increase, so a noticeable rise in this tracing after a relevant question is flagged as a potential indicator of deception.
How Examiners Score the Results
Scoring happens either by hand or with computer algorithms, and both methods follow the same basic principle: for each relevant question, the examiner compares the physiological response to the response seen after the nearest control question. Each channel (breathing, sweat, cardiovascular) is evaluated separately.
In manual scoring, examiners assign a numerical value to each comparison, typically on a scale that ranges from negative numbers (indicating a stronger response to the relevant question, suggesting deception) to positive numbers (indicating a stronger response to the control question, suggesting truthfulness). A score of zero means no meaningful difference was observed. These scores are then added across all channels and all question repetitions to produce a grand total. A sufficiently negative total leads to a “Deception Indicated” result, a sufficiently positive total means “No Deception Indicated,” and scores in between fall into an “Inconclusive” category.
Computerized scoring systems extract the same features, measuring amplitude changes, response duration, rate shifts, and baseline movement, but apply statistical algorithms rather than human judgment. Both methods are looking at the same data; they just differ in how consistently they weigh each feature.
Why the Results Are Contested
The fundamental problem with polygraph interpretation is that it doesn’t detect lies. It detects physiological arousal, and arousal can be triggered by anxiety, anger, confusion, or simply being nervous about being accused. A truthful person who feels frightened by a relevant question can produce the same spike as someone who is lying. A meta-analysis synthesizing results from 206 studies found that people (including trained evaluators) achieve only about 54% accuracy in distinguishing lies from truth, with 47% of actual lies correctly identified and 39% of truthful statements incorrectly flagged as deceptive.
The National Academies of Sciences reviewed polygraph research and identified the core issue: the same autonomic nervous system responses that increase blood pressure, activate sweat glands, and alter breathing rate during deception also fire during any form of emotional stress. There is no unique physiological signature for lying.
Countermeasures and Their Limits
Because the test relies on comparing responses to control versus relevant questions, a knowledgeable subject can attempt to game the system by artificially boosting their reaction to control questions. Techniques like clenching a toe, biting the tongue, or doing mental arithmetic during control questions can create spikes that make the relevant-question responses look smaller by comparison. Research from the Department of Justice confirms that a sophisticated subject can generate control-question reactions that exceed their responses to relevant items, potentially producing a false “truthful” result.
Examiners are trained to watch for signs of countermeasures, including unusual movements, overly regular breathing patterns, or inconsistent physiological data that doesn’t match what they observe visually. Some modern setups include motion-sensing pads in the exam chair to detect subtle physical manipulations.
Legal Status of Polygraph Results
Most U.S. courts do not admit polygraph results as evidence. There is no federal statute or rule of evidence that specifically addresses polygraphs, and for decades the majority of federal and state courts treated them as inadmissible because the technique lacks general acceptance in the scientific community. A 1993 Supreme Court decision (Daubert v. Merrell Dow Pharmaceuticals) loosened the standard by ruling that scientific evidence shouldn’t be excluded solely for lacking general acceptance, but judges must still evaluate whether the technique has been properly tested, peer-reviewed, and whether it has a high error rate. Most courts that have applied these criteria continue to exclude polygraph evidence, often noting that the results could mislead juries or waste the court’s time.
Some jurisdictions allow polygraph results when both parties agree in advance to their admissibility, and a small number of federal circuits have moved away from blanket exclusion. But as a practical matter, polygraph results carry far more weight in employment screenings, security clearances, and law enforcement investigations than they do in courtrooms.