A $2 Drug Test That Gets It Wrong Up to 38% of the Time — and Why Georgia Must Act

TL;DR

Colorado just passed the first law in the country to stop arrests based only on cheap roadside drug tests. These $2 test kits get it wrong up to 38% of the time. They have led to about 30,000 false arrests each year across the U.S. Georgia is the only state where these bad test results can be used to convict someone at trial. About 961 Georgians are falsely arrested each year because of these tests.

Why This Matters

If your loved one is in a Georgia jail on a drug charge, there’s a real chance the drug test was wrong. These cheap tests have flagged cotton candy, donut glaze, and baby powder as drugs.

In Georgia, police can arrest someone based on just one of these tests. Then the state can convict them based on that same bad test — without ever doing a real lab test. No other state allows this.

Most people can’t afford bail. They sit in jail for weeks or months. They lose jobs. They miss births, funerals, and family crises. Many plead guilty just to get out — even if they are innocent. Once they plead guilty, the evidence is almost never tested in a real lab. The truth stays hidden.

This means families lose loved ones to prison based on a test that is “only slightly better than a coin flip,” as one judge put it.

Key Takeaway: Georgia is the only state where a $2 roadside drug test can convict someone at trial — no lab test needed.

What Colorado Did

On March 26, 2026, Colorado’s governor signed a new law called HB 26-1020. It passed with zero “no” votes — 65-0 in the House and 33-0 in the Senate.

The law does two main things:

• Stops arrests based only on a field drug test. For lower-level drug charges, police must give a court date (called a summons) instead of making an arrest.
• Tells people their rights before a guilty plea. Courts must warn people that field tests often get it wrong, that field test results can’t be used in court, and that they have the right to ask for a real lab test.

The law costs the state nothing. It cuts jail costs while adding only a little more work for courts.

The push for this law started after Holly Bennett’s case. Bennett is a 65-year-old great-grandmother. Police charged her with cocaine while she was in a hospital bed — for neck surgery. The “cocaine” was her crushed-up prescription Ritalin. She spent 15 months fighting the charge. She had to take out a new loan on her home to pay for a lawyer.

Key Takeaway: Colorado’s law passed with zero opposition and costs the state nothing.

How These Tests Fail

A major 2024 study from the University of Pennsylvania looked at field drug tests across the whole country. Here’s what they found:

• The makers of these tests claim they are wrong only about 4% of the time.
• The real error rate is between 15% and 38% — up to nearly ten times higher.
• About 773,000 drug arrests each year use these tests. That’s half of all drug arrests.
• About 30,000 people are falsely arrested each year because of bad results. The researchers said this number is likely too low.

The study called field drug testing “one of the largest known factors in wrongful arrests” in the country.

Why do these tests fail so often?

The tests look for groups of chemicals, not specific drugs. Many legal items share the same chemical traits as drugs. For example:

• The test for cocaine also reacts to lidocaine and cold medicine.
• The test for meth also reacts to sugar.

The tests also depend on a person reading a color change. Lighting, eyesight, and bias all affect the reading.

Things that have tested “positive” for drugs:

• Cotton candy → meth
• Donut glaze → meth
• Bird droppings → cocaine
• Powdered milk → cocaine
• A toddler’s ashes → MDMA/meth
• Stress ball sand → cocaine
• IBS medicine → fentanyl
• Kitty litter → meth

In New York City jails, 71 items that tested positive for fentanyl were sent to a real lab. Only 15% actually had fentanyl. That’s an 85% false-positive rate. One brand of test — the NARK II — had a 91% false-positive rate.

No federal agency checks or controls how these tests are made or sold.

Key Takeaway: Field drug tests get it wrong up to 38% of the time — and common items like sugar, donut glaze, and cold medicine can trigger a false positive.

The Wrongful Conviction Crisis

Houston, Texas: The largest known scandal

Between 2004 and 2015, at least 298 people were convicted of drug crimes in Harris County, Texas — even though real lab tests later found no drugs at all. Of those, 212 were based on field tests done by Houston police.

The county’s review unit overturned more than 131 of these cases. Harris County was behind half of all overturned convictions by such units in the whole country.

The innocent people pleaded guilty an average of 4 days after arrest — long before any lab could check the evidence.

Amy Albritton’s story shows how this works. In 2010, police in Houston found a white crumb on her car floor. A $1 test said it was crack cocaine. She pleaded guilty within 48 hours. She served 21 days in jail. Five months later, a real lab found no drugs. But no one told her for years. Her record wasn’t cleared until 6 years after her false arrest.

Las Vegas, Nevada

In Las Vegas, 33% of cocaine field tests between 2010 and 2013 were wrong. In 2014, only 8 of 4,633 drug cases went to trial. The rest — 99.8% — were plea deals.

Georgia’s own crisis

In 2017, the GBI Crime Lab found 145 false positives from field tests in Georgia in just one year:
– 64 for meth
– 40 for cocaine
– 24 for ecstasy
– 11 for heroin

At least three people had already pleaded guilty before the lab results came back.

A review by Savannah police that same year found the test was wrong in 9 out of 42 cases — a 21.4% error rate.

In New York, the State Inspector General found that 2,000 people in prison were wrongly punished based on bad field tests. Punishments included solitary, lost visits, and lost good-time credits.

Key Takeaway: At least 298 people were convicted in one Texas county alone when lab tests showed no drugs were present.

Georgia Stands Alone

Georgia is the only state where field drug test results can be used to convict someone at trial. This comes from court rulings, not a law that voters or lawmakers chose.

In 2006, a Georgia court ruled that a positive field test alone is “enough to sustain a conviction.” No real lab test is needed. No other state allows this.

How Georgia police use these tests

A 2024 survey of Georgia police chiefs found:
– 68% of agencies still use roadside drug tests.
– Only 26% have a rule that says officers need more proof before making an arrest.

There is no required training for officers on how to use these tests.

Dasha Fincher’s story

On New Year’s Eve 2016, deputies in Monroe County pulled over a car. They found a clear bag with a blue substance on the floor. It was cotton candy.

A field test said it was meth. Dasha Fincher was charged with drug trafficking. Her bail was set at $1 million cash.

She spent 94 days in jail. While locked up, she missed the birth of her twin grandchildren. She missed her daughter’s miscarriage. She was denied care for a broken hand.

The GBI lab confirmed there were no drugs — but it took over two months. Even after the lab cleared her, charges weren’t dropped for almost another month.

Fincher sued. The court threw out her case. The judge wrote that she “should never have spent 94 days in jail” — but said the law protected the deputies. She got nothing.

The GBI lab backlog

Georgia’s state crime lab tests drugs for about 800 police agencies across 159 counties. In 2019, the lab had a backlog of 36,194 items. Over half — 19,112 — were drug tests.

The lab has made progress. In 2024, it finished about 105,000 tests while getting about 103,000 new requests. The backlog dropped by 11%. But results still take weeks to months — and that gap is when innocent people plead guilty.

Key Takeaway: About 961 Georgians are falsely arrested each year due to faulty field drug tests, and Georgia is the only state where those results can convict someone.

Guilty Pleas Hide the Truth

This is where the biggest harm happens — and where it stays hidden.

About 95% of all criminal cases end in plea deals. People plead guilty without ever going to trial. Here is what studies found about drug cases:

• 89% of prosecutors accept guilty pleas with no real lab test.
• 67% of drug labs say they are never asked to test samples when a case ends in a plea deal.
• 46% of labs will not even run a test if someone already pleaded guilty.
• Over 80% of prosecutors said it is “very unlikely” that drug evidence will ever be tested once a deal is made.

At least 100,000 people each year plead guilty to drug charges based mainly on field test results.

Think about what this means. If the evidence is never tested, no one ever finds out the test was wrong. The false conviction stays on the books. The person has a record. They lose housing, jobs, and custody rights — for a crime that never happened.

The only reason we know about the 298 false convictions in Houston is that the local lab tested evidence even after guilty pleas. Most places don’t do this.

Cody Gregg was a 26-year-old homeless man. Police arrested him for powdered milk that tested positive for cocaine. He pleaded guilty to drug charges and was sentenced to 15 years in prison. He just wanted to get out of the crowded jail. The lab found no drugs nearly two months later. His plea was taken back and the case was dropped.

The National Registry of Exonerations has records of 3,396 cases where people were cleared of crimes. Of those, 531 involved drug arrests for items that were not drugs.

Key Takeaway: 89% of prosecutors accept guilty pleas without any real lab test — meaning thousands of false convictions are never discovered.

Race and Poverty Decide Who Suffers Most

The harm from bad drug tests does not fall equally.

Race

The 2024 Penn study found that Black Americans are falsely arrested from field tests at three times the rate of white Americans.

In Harris County, Texas, 60% of people wrongly convicted were Black — in a city that is only 24% Black.

Of people wrongly convicted in cases studied by ProPublica, 93% received jail or prison time.

Poverty

The system punishes people who can’t afford to fight back.

• Dasha Fincher could not pay her $1 million bail for cotton candy. She sat in jail for 94 days.
• Kena’z Edwards could not pay his $178,000 bail. His lidocaine tested as cocaine. He sat in jail for over three months.
• Cody Gregg pleaded guilty to a 15-year sentence for powdered milk just to escape jail.
• Holly Bennett had to refinance her home to pay a lawyer for 15 months.

A person with money can post bail and wait for lab results to clear them. A person without money stays locked up. The pressure to plead guilty becomes crushing.

The Penn study noted that these problems stack on top of racial gaps that already exist. Black and Hispanic people are already stopped and searched more often — even though drug use rates are similar across races.

Key Takeaway: Black Americans are falsely arrested from field drug tests at three times the rate of white Americans.

What Georgia Should Do Next

Colorado has shown a clear path. Georgia can follow it — but must go further because of its harsher drug laws.

In Georgia, having any amount of cocaine, heroin, fentanyl, or meth is a felony. Penalties range from 2 to 15 years for simple possession. Colorado’s law mostly covers lesser charges. A Georgia law must cover felonies too.

A Georgia reform bill should do five things:

1. Stop arrests based only on field tests. Officers should give a court date instead of making an arrest when the only evidence is a field test.
2. Require courts to warn people before a guilty plea. Tell them the test is often wrong. Tell them they can plead not guilty. Tell them they can ask for a real lab test.
3. Ban field test results from being used as evidence at trial. This would overturn the 2006 court ruling that lets Georgia convict people based on these tests alone.
4. Give everyone the right to a real lab test from the GBI or another certified lab at any point in their case.
5. Let people take back a guilty plea if a lab test later shows there were no drugs.

Georgia has a history of criminal justice reform. Governor Kemp signed a law in 2025 paying wrongly convicted people $75,000 per year. The Georgia Innocence Project says 44% of its cases involve bad forensic evidence.

Colorado’s law cost $0 in new spending. It had support from groups on both the left and the right. Georgia can do the same.

Key Takeaway: Georgia needs a law that bans field test results from trial, stops test-only arrests, and gives everyone the right to a real lab test.

Glossary

• Field drug test (colorimetric test): A cheap chemical kit (about $2) used by police on the roadside. You crack a tube and watch for a color change. It is meant to give a quick guess — not a final answer.

• False positive: When a test says “drugs” but there are no drugs. Many legal items — sugar, cold medicine, donut glaze — can cause this.

• Lab test (confirmatory testing): A real test done in a lab using advanced tools. It can tell exactly what a substance is. It takes weeks to months.

• Plea deal (plea bargain): When a person agrees to plead guilty — often to a lesser charge — instead of going to trial. About 95% of cases end this way.

• Summons: A paper that orders someone to come to court on a set date. It replaces an arrest.

• GBI Crime Lab: Georgia Bureau of Investigation’s state lab. It tests drug evidence for about 800 police agencies across the state.

• Harper standard: Georgia’s legal test for whether scientific evidence can be used in court. Courts have said field drug tests meet this test — even though the science says otherwise.

• Exoneration: When a wrongly convicted person is officially cleared of a crime.

• Sovereign immunity: A legal rule that protects government workers from lawsuits. It was used to block Dasha Fincher’s case even though the judge agreed she should never have been jailed.

• Backlog: The pile of untested evidence waiting at the lab. In 2019, Georgia had over 36,000 items waiting — more than half were drug tests.

Read the Source Document

Read the full GPS policy brief: “Colorado’s landmark drug test law and what it means for Georgia” (PDF)

Other Versions

We created versions of this post for different audiences:

• For Legislators — Policy details, bill language, and fiscal impact
• For Media — Key findings, data points, and case studies
• For Advocates — Talking points, coalition info, and action steps