The laboratory can say a driver had no alcohol or drugs in their system. It just cannot say it soon enough to prevent the handcuffs, the tow bill, the night in jail or the public arrest record.
That is the failure exposed by a WSMV4 investigation, recently republished by WIS, that identified cases in 22 states where drivers were arrested for DUI despite later evidence showing no intoxicants. The 22-state figure is the newsroom’s finding from police reports, lawsuits and news coverage, not an official federal count.
North Carolina law gives officers many of the same tools used in those cases: roadside sobriety tests, alcohol screening and blood or urine analysis when drugs are suspected. The state crime lab processes thousands of toxicology submissions, but Killjoy Media could not find a published statewide count of DWI arrests in which testing detected neither alcohol nor drugs.
That means North Carolina cannot publicly answer the most basic question raised by the national investigation: How often does the evidence clear a person only after the system has already treated them as guilty?
The arrest happens before the laboratory answers
One case documented by WSMV involved Justin Beery, who was stopped in Pelham, Alabama, in March 2025 after working a 14-hour shift on roughly three hours of sleep.
According to the station’s reporting, Beery told officers he was autistic and struggled through field-sobriety exercises. An officer arrested him despite reporting no odor of alcohol and finding no alcohol containers or drugs. A later blood test detected neither alcohol nor drugs, and the charge was dropped. The Pelham Police Department opened an internal investigation and declined further comment to WSMV while it was underway.
Beery’s case is not proof that every questionable DWI arrest involves bad faith. It is proof that behavior an officer reads as impairment can have another explanation—and that the definitive evidence may not exist until after the arrest.
That sequence matters. Probable cause is a lower standard than proof beyond a reasonable doubt. An officer does not need a finished toxicology report to make an arrest. But when the suspicion is wrong, the driver’s first reliable scientific answer arrives after the most coercive parts of the process have already happened.
Tennessee counted 2,547 cases only after someone forced the question
Tennessee provides the clearest view of the scale because its Bureau of Investigation has data—and because reporters kept asking for it.
In January, the TBI corrected an earlier response that had counted 609 DUI arrests from 2017 through 2023 with no alcohol or drugs detected. The corrected number was 2,218. Another 419 cases in 2024 brought the total to 2,547.
The annual count fell to 240 in 2020 and then climbed to 419 in 2024. The Tennessee Highway Patrol accounted for 180 of the 2024 cases, according to the data reported by WSMV.
The TBI disputes calling every one of those drivers “sober.” That caveat is legitimate: toxicology panels do not test for every conceivable impairing substance, and a negative report cannot prove what a laboratory never looked for. A clean toxicology result also does not automatically establish that an officer lacked probable cause at the roadside.
But the caveat does not make 2,547 disappear. It makes the next questions more important: Which substances were tested? Why was each person arrested? What did body-camera footage show? Which officers repeatedly made arrests followed by negative results? WSMV reported that the charges were dismissed in every case it examined where neither alcohol nor drugs were detected.
Tennessee lawmakers responded with a 2025 law requiring the TBI to report annually through 2029 on DUI arrests with no detected intoxicants and to identify the arresting agencies. That is not radical reform. It is basic quality control.
Field sobriety tests can detect impairment—and still accuse sober people
Standardized field-sobriety tests were developed and validated primarily around alcohol impairment. The National Highway Traffic Safety Administration says the three-test battery can identify drivers at or above a 0.08 blood-alcohol concentration with about 91% accuracy when properly administered.
That is useful evidence. It is not magic, and it is not the same as a drug test.
A 2023 randomized clinical trial published in JAMA Psychiatry examined whether highly trained officers could identify cannabis impairment. Officers classified 81% of participants who received active THC as impaired—but they also classified 49.2% of the placebo group as impaired.
The study did not conclude that field tests are worthless. Some test components correlated with simulated-driving performance, and officers distinguished the active group from the placebo group better than chance. Its warning was narrower and more important: without other indicators, field tests may be insufficient to identify impairment specifically caused by THC.
In plain English, a roadside exercise can contribute to a reasonable investigation without proving why a person appears unsteady, slow, confused or anxious. Fatigue, disability, injury, medication, age, stress and simple difficulty following instructions can all complicate what an officer sees.
This is not only a Tennessee problem
WSMV found reported cases in 22 states, including Alabama, Florida, Hawaii and Iowa. The exact facts differ, but the pattern is familiar: an officer observes driving or behavior believed to show impairment, an alcohol test is negative or near zero, suspicion shifts to drugs, and laboratory results later fail to confirm it.
In Honolulu, a federal lawsuit filed by the American Civil Liberties Union alleges that police arrested 127 people between 2022 and 2024 after breath or blood alcohol tests showed 0.000. According to Associated Press reporting on the lawsuit, only 15 received drug-related citations and three were charged with drug-impaired driving.
Those are allegations, not final findings. Honolulu police said the department was reviewing impaired-driving arrests dating to 2021 and took the claims seriously. The lawsuit also alleges that officers had incentives to make DUI arrests, including the ability in some circumstances to finish a shift early while still being paid. Police accountability requires examining that allegation, not pretending incentives never influence human behavior because a policy manual uses the word “integrity.”
Could it happen in North Carolina?
Yes—not because every North Carolina DWI arrest is suspect, but because the legal and scientific sequence makes a mistaken arrest possible here too.
Under G.S. 20-138.1, a person can be charged with impaired driving based on being under the influence of an impairing substance, having a blood-alcohol concentration of at least 0.08 or having any amount of a Schedule I controlled substance or its metabolites in blood or urine.
North Carolina’s roadside alcohol-screening law explicitly says a negative screening result may be used in an appropriate case to help determine whether suspected impairment came from something other than alcohol. A zero does not necessarily end the investigation. It can redirect it.
The North Carolina State Crime Laboratory uses mass-spectrometry and other methods to examine blood and urine for alcohol, inhalants, prescription drugs and illicit drugs. Its current evidence-submission policy says blood is preferred in DWI cases and instructs agencies to identify suspected substances when requesting drug analysis.
That process can confirm an officer’s suspicion. It can also fail to confirm it. What the public cannot see is how often that second outcome follows an arrest.
Killjoy found annual crime-lab workload figures and testing policies, but no public North Carolina report comparable to Tennessee’s agency-by-agency count of DWI arrests followed by toxicology showing no alcohol or drugs. That does not prove the state lacks the data. It means the state does not appear to publish the answer in a form residents, lawmakers or reporters can readily evaluate.
North Carolina should count negative toxicology cases
A negative laboratory result does not, by itself, prove misconduct. Some substances fall outside routine panels. Evidence available during a stop can differ from evidence available later. Prosecutors may dismiss a case for several reasons.
Those limits are arguments for better data, not excuses for having none.
North Carolina should publish the number of DWI arrests followed by reports detecting no alcohol or tested drugs, broken down by arresting agency and year. Agencies should audit officers with repeated negative-result cases, review body-camera footage, document medical or disability explanations raised during stops and disclose how quickly prosecutors are notified when testing fails to support a charge.
The state should also report what each toxicology panel includes. Calling a result “negative” without explaining the panel invites overstatement; refusing to count negative results because no panel tests for everything invites institutional blindness.
This is not a demand to stop enforcing impaired-driving laws. Drunk and drug-impaired drivers kill people. It is a demand that enforcement systems measure their own errors with the same enthusiasm they bring to measuring arrests.
North Carolina already counts DWI arrests. Counting the people the laboratory fails to incriminate should not be the complicated part.
Killjoy Media reviewed reporting from WSMV, WIS and The Associated Press; a peer-reviewed randomized clinical trial; NHTSA materials; North Carolina statutes; and North Carolina State Crime Laboratory policies. Continue reading Killjoy’s police accountability coverage.