How Police Radar Works: Doppler, Modes and Checks

By I Go Cars · Updated October 29, 2026

What police radar is and what it measures

A police radar unit, known formally as a radar speed gun, measures how fast a vehicle is moving by transmitting a microwave beam and measuring how the frequency of the reflected signal has changed. That change is the Doppler shift. The National Highway Traffic Safety Administration (NHTSA), the federal agency that writes the model training curriculum for speed-measuring devices, defines the principle in its training glossary: "When there is relative motion between two objects, one of which is transmitting wave energy," the frequency received differs from the frequency sent. The size of that difference is proportional to the relative speed.

NHTSA's model course is called Speed-Measuring Device (SMD) Operator Training. It has a Core module covering legal aspects, speed estimation and tracking history, plus device modules for radar and lidar. The radar module's instructor manual (September 2018) and the SMD Glossary (NHTSA's list of defined training terms) are the primary documents behind this page. NHTSA writes the device name as "R.A.D.A.R." throughout, an acronym for radio detection and ranging.

The direction of the shift tells the device which way the target is moving. In the manual's words, "If the target is moving toward the R.A.D.A.R. device, the frequency is increased." A receding target lowers the returned frequency. A beam that strikes nothing returns nothing: "A R.A.D.A.R. beam transmitted into empty space will go on forever and obtain no reading." Police radar does not photograph, identify or track a specific car. It reports a number, and the officer decides which vehicle produced it.

Stationary mode: the basic measurement

Stationary mode is, in the SMD Glossary's definition, "the capability of a R.A.D.A.R. device to operate from a fixed location and display" a target's speed. The patrol car is parked; only the target moves. The returned signal's Doppler shift converts directly into a reading, and the counting unit (the box that processes the signal and drives the display) shows it.

Stationary mode is the simplest case because there is only one moving object in the calculation. The main geometric error, the cosine effect, can only lower the reading in this mode. NHTSA's manual says the stationary cosine effect "will always be to the motorist's advantage," because the radar sees only the part of the target's motion that is directly toward or away from the antenna.

Moving mode: adding the patrol car's own speed

NHTSA's manual defines it plainly: "Moving R.A.D.A.R. is a R.A.D.A.R. device being operated from a moving patrol vehicle." Moving capability arrived in 1972, according to the same manual. In moving mode the device takes two readings at once: a low-frequency return from the road and roadside (the patrol speed) and a return from the target vehicle (the closing or separating speed). It then subtracts or adds to calculate the target's own speed.

Opposite-direction targets

For an oncoming car seen through the front antenna, the radar measures the closing rate (the two vehicles' velocities added together) and subtracts the patrol figure. If the patrol car is doing 50 mph and the two are closing at 110 mph, the display shows 60 mph. NHTSA's opposite-lane tracking history includes an explicit step: "Verify Patrol Speed."

Same-direction targets

Same-direction moving radar is harder. NHTSA states that "Normally, same-direction moving R.A.D.A.R. devices are unable to automatically differentiate" whether the target is faster or slower than the patrol car, and "The operator must determine which case is present and manually switch the R.A.D.A.R." That is a human-judgment step built into the measurement. The same-lane tracking history adds a speed discrimination test for this reason.

The practical point is simple. Moving mode is only as good as its patrol-speed reading. Every moving-mode error NHTSA describes traces back to that second number.

The beam, its range and the strength of the signal

A radar beam spreads as it travels. NHTSA's manual says "The angle of the emitted R.A.D.A.R. beam will determine the relative beam width," and gives a range of 9 to 18 degrees depending on the manufacturer. An 18-degree beam is about 80 feet wide at 250 feet, about 160 feet wide at 500 feet and about 320 feet wide at 1,000 feet. At those distances the beam can cover several lanes and several vehicles at once, which is why the operator's visual observation matters so much.

Signal strength falls quickly with distance. The manual's inverse-square example puts the return signal 16 times weaker at 1,000 feet than at 250 feet. NHTSA states no maximum range in feet; range depends on the target's size and shape, the device and the conditions. Weather has a modest effect: "Rain or snow will decrease the R.A.D.A.R.'s range slightly."

Radar units also produce an audio Doppler tone, and NHTSA trains operators to listen to it: "A strong, clear tone indicates a good, clear, received signal," while "A broken or scratchy audio tone indicates a weak received signal."

Police radar operates in three frequency bands that NHTSA defines in its performance specification: X-band (10.500 to 10.550 GHz), K-band (24.050 to 24.250 GHz) and Ka-band (33.4 to 36.0 GHz). The band changes the frequency, not the principle; the Doppler measurement works the same way in all three.

How accurate police radar is supposed to be

NHTSA's Down-the-Road Radar Module specification (DOT HS 812 266), the performance standard that radar models were tested against, requires stationary-mode readings within +1/-2 mph (+2/-3 km/h) across 20 to 100 mph, and moving-mode patrol and target speeds within ±2 mph (±3 km/h). The asymmetric stationary tolerance means a compliant unit may read up to 2 mph low but no more than 1 mph high. The wider moving-mode tolerance reflects the extra patrol-speed calculation.

Models that passed independent testing appeared on NHTSA's Conforming Product List (CPL). The November 28, 2022 CPL listed mostly Ka-band down-the-road radars, with a few K/Ka units and one X/K/Ka unit, the MPH Python III. In November 2024 NHTSA announced in the Federal Register that it would discontinue the radar CPL, keep it for one year, and shift to an industry-based verification program built on IEEE Standard 2450. That transition window ended around November 2025; the current status of the replacement program should be checked before relying on it. The specification remains the published NHTSA benchmark, not a current federal mandate.

Daily checks

NHTSA training relies on two checks before and after use:

Internal circuit test: it "checks the circuits inside the counting unit." It does not check the antenna.
Tuning-fork test: a tuning fork is a calibrated metal fork that, when struck and held in front of the antenna, produces a known Doppler signal. "Traditionally, only one tuning fork is needed to check a stationary R.A.D.A.R. device"; "For moving R.A.D.A.R., two forks are necessary." The reading may not differ from the fork's certified value by more than 1 mph.

"Agencies may require these tests be performed at least twice, at the beginning and end" of a shift. NHTSA also tells operators that automatic lock and alarm features are "not approved" and should never be used: the operator, not the device, decides when to lock a reading.

Error sources NHTSA trains operators to recognize

NHTSA's manual is direct about what can go wrong. The errors fall into two groups: those specific to moving mode and those that affect any radar.

Moving-mode errors

Shadowing: the radar takes its patrol speed from a slower vehicle traveling in the same direction instead of from the road, so the patrol figure reads low and the target figure reads high. The operator catches it by comparing the displayed patrol figure with the speedometer.
Batching: "Spurious or no target speed display readings due to rapid changes in patrol vehicle speed." The manual states that current devices are not subject to it.
Moving cosine effect: an angle between the radar and the target, or between the radar and the roadside it uses as its own-vehicle reference. NHTSA says it "most often" produces a low reading, but it can produce a high one when patrol speed is under-read.

Interference in any mode

Moving signs, roof air-conditioning fans and other mechanical devices.
Radio-frequency interference (RFI) from radios inside or near the patrol car.
Mercury vapor, neon and fluorescent lighting.
The panning effect: a handheld antenna swung past a parked car or other object, producing a false reading.

None of these errors is mysterious, and none is hidden from officers. They are taught as named conditions with named checks. The defense-relevant question in a moving-mode case is usually whether the officer verified patrol speed, which NHTSA's tracking history requires.

The operator, the tracking history and the visual estimate

A common belief is that the radar number is the whole case. NHTSA training says otherwise: "The visual element is the first part of any speed enforcement action." The Core module treats speed estimation as a trained skill ("The ability to estimate speed is a learned skill") and teaches a tracking history in three parts: visual observation and estimation, audio confirmation, and unit confirmation. The radar reading confirms what the officer has already seen and heard.

The Core manual also gives a working figure: "A vehicle will travel approximately 1.5 feet during each second for each mile per hour of speed." A car at 60 mph covers about 88 feet each second. It also notes that "Approximately two thirds of the States have absolute speed limits," while others use a basic speed law requiring a "reasonable and prudent speed for existing conditions." Whether an officer's visual estimate alone supports a conviction is a matter of state law that these sources do not answer.

Operator certification is set state by state; this page does not survey it. On military installations, federal rules (32 CFR 634.27) require traffic radar operators to meet the host state's training and certification requirements and to be recertified every three years, or as state law prescribes. The training manuals themselves are published on the NHTSA radar instructor manual page.

Where police radar came from

NHTSA's manual traces the chain: Christian Doppler described the frequency effect in 1842; Heinrich Hertz observed radio waves in 1886; radar was developed for detecting military aircraft in 1935; and "In 1947, law enforcement began using R.A.D.A.R. for measuring the speed of vehicles." A November 2024 Federal Register notice from NHTSA describes radar as used to detect speeding motorists since the late 1940s, consistent with the 1947 date, and records that NHTSA began working with the National Bureau of Standards on radar standards in 1977. Moving radar followed in 1972.

Misconceptions NHTSA itself lists

The manual names several public beliefs as misconceptions: that aluminum foil, chains, hubcaps or honking the horn defeat radar. They do not. It also corrects the belief that moving radar always reads high; NHTSA says the moving cosine effect most often reads low, and a high reading requires an under-read patrol speed that a careful operator can see on the display.

Frequently asked questions

How does police radar measure speed?

Police radar measures a vehicle's velocity by transmitting a microwave beam and measuring the Doppler shift in the reflected signal. An approaching vehicle raises the returned frequency and a receding one lowers it; the size of the shift is proportional to speed. NHTSA's 2018 radar instructor manual describes the principle, and the counting unit converts the shift into a displayed reading.

Can police radar work while the patrol car is moving?

Yes. Moving radar has existed since 1972, according to NHTSA's training manual. The device measures how fast the patrol car itself is moving from the road and the closing or separating rate of the target, then calculates the target's velocity. NHTSA requires the officer to verify the displayed patrol speed against the speedometer, because an under-read patrol figure produces a high target reading.

How accurate is police radar?

NHTSA's down-the-road radar specification requires stationary readings within +1/-2 mph between 20 and 100 mph, and moving-mode patrol and target speeds within plus or minus 2 mph. Those are performance standards a device model is tested against. In use, accuracy also depends on the operator's checks, setup and target identification, which NHTSA training covers in detail.

How do officers check that a radar unit is accurate?

Officers run an internal circuit test and a tuning-fork test. NHTSA training uses one fork for stationary radar and two for moving radar, and the reading may not differ from the fork's certified value by more than 1 mph. Agencies may require the checks at least twice per shift, at the beginning and the end.

When did police start using radar?

Police in the United States began using radar for vehicle speed in 1947, according to NHTSA's radar training manual. A 2024 NHTSA notice in the Federal Register says radar has been used to detect speeding motorists since the late 1940s, which is consistent. NHTSA began work on radar standards with the National Bureau of Standards in 1977.

Can police give a speeding ticket without radar?

NHTSA training treats an officer's visual speed estimate as a learned skill and the first step of every speed stop, with radar serving as confirmation. Whether a visual estimate alone is enough to support a conviction is a matter of state law, which varies. Check your state's statutes and court decisions for the current position.

Does aluminum foil or a hubcap block police radar?

No. NHTSA's radar instructor manual lists foil, chains, hubcaps and honking the horn as public misconceptions about defeating radar. A vehicle's body reflects the microwave signal regardless of these items, and the Doppler measurement is unaffected by them. The same manual lists real error sources, such as shadowing and radio interference, which officers are trained to recognize.

How police radar works