Hillsborough County deputies arrested 15 people on drug charges in the first months of 2014 using chemical field tests, and the Florida state crime lab later found that none of those substances were drugs. That’s from ProPublica’s reporting with the Tampa Bay Times. A field test is a cheap pouch of chemicals: an officer drops in the suspected substance and watches for a color.
Amid those mistakes, a sheriff’s lieutenant ran a test of the kits, saw an empty pouch turn the same color as a “positive” one, and wrote a memo. Deputies were ordered to turn in their kits, and prosecutors dropped a number of methamphetamine possession charges. Sheriff’s officials then realized the lieutenant had misread the colors, and the reporters wrote that little could be done about meth cases that may have been dismissed too early. The sheriff’s office didn’t change its approach to training afterward and switched instead to another meth kit from the same maker, one whose reagent (the chemical inside the pouch that changes color) turns the same orange-to-brown for sugar as it does for meth.
The same reporting found that in 2014, one in four Hillsborough submissions described as methamphetamine came back from the state lab as something else, and that statewide in 2015, 21 percent of material suspected of being meth wasn’t. The reporters couldn’t say how many of those submissions involved a field test. Either way, those numbers exist only because a lab checked.
What a 2026 Strip Study Counted
A study published in Clinical Toxicology in 2026, by Chun, Mattu, Lee and Wilkerson, did that checking on purpose for a different quick test. It took 250 leftover urine specimens from hospital patients and ran each one through a commercial test strip for xylazine (a sedative the FDA has approved for animals, not people, now turning up in the street supply alongside fentanyl) and then through liquid chromatography with mass spectrometry, the lab method that names the exact chemical.
The instrument found xylazine in 71 specimens. The strip caught 19 of them, a sensitivity of 26.8%, and missed 52. It turned positive 59 times, and 40 of those were wrong, so a positive strip was right about one time in three (the study’s positive predictive value is 32.2%). The authors also report that 91.5% of the real positives held less xylazine than the 1,000 ng/mL the manufacturer lists as the strip’s threshold. A test can work exactly as designed and still miss when the real amounts sit below what it was built to see. Even the fentanyl screen that got each specimen into the study was unconfirmed.
What the Study Doesn’t Show
It doesn’t measure the reagent pouches officers carry, and it doesn’t test powders, pills or residue. The strips were made for drug samples and used here on urine, and the authors put the poor showing down largely to low sensitivity at the levels real patients had. It’s one hospital and one commercial strip. Anyone who quotes these numbers as the error rate of a roadside cocaine test is misreading it.
What it does show is what a fast result is worth before anyone checks it against a specific lab method on the same sample: not much. My screening vs confirmation page walks through why a presumptive result is a lead and a confirmation is where identity comes from.
A Presumptive Test Leaves Paperwork Too
A color test or a strip is still a test, and it leaves records. When a lab or an agency runs a preliminary test, whether a color test, a crystal test or an immunoassay screen (an antibody test that reacts to a drug class), there should be a written procedure for it, a lot number for the reagent, quality control results for that lot, and a record of what was seen, when, and by whom. If a positive strip or pouch helped justify the arrest, I want to know whether anyone checked that reagent lot against a known sample, and what the procedure says a positive looks like.
Then the confirmation. A lab report that names a drug rests on a validated method, and the validation should cover how the method behaves in the kind of sample tested, which is the exact problem the xylazine study ran into. When a screen fed the lab’s decision, I ask for the cross-reactivity data for that screen, meaning the list of other substances that can set it off, and the cutoff the lab applied. Having a written procedure doesn’t prove the procedure is any good, and the validation file is where that question gets answered.
What I Ask For in a Florida Case
Florida’s criminal discovery rule, Rule 3.220, requires the State to disclose reports or statements of experts made in connection with the case, including the results of scientific tests, and to let the defense inspect, copy, test and photograph material in its possession. When the reports alone don’t answer the question, the rule lets the court order additional discovery when the defense shows the material matters to the case. In a drug case, I send a specific request list rather than a one-line demand for the lab report.
From the field, I ask for the brand and lot of any test kit or strip, photographs of the result, and the officer’s training on that kit. From the lab, I ask for the procedure for each screening and confirming method, the validation studies for those methods, the chromatograms and spectra (the instrument’s printouts that show what it actually detected) with the run log, and the quality control results for the batch. I also ask for the analyst’s training records and proficiency tests, and the full chain of custody from the scene to the bench. I want the sample itself available for independent testing too, because the rule allows it.
None of that assumes anyone did anything wrong. A presumptive result and a confirmed one are different claims, and the paperwork shows which one the case actually has.
When the Case Started With a Color Change
If your arrest started with a pouch turning blue or a strip showing a line, that result may have helped support the arrest, but it’s a presumptive call, and what the substance was still has to be proven. The question is whether a lab later confirmed it with a specific method, whether that method was validated for the sample it was used on, and whether the records back the result up. A study like this one is a reminder that quick tests get counted wrong in both directions, and that the count only shows up when somebody checks.
My drug possession defense page covers how possession cases are built in Florida, and challenging the drug evidence covers the lab side. If the charge is a drug DUI, the drug testing in a drug DUI section lays out how blood and urine results are screened, confirmed and read.
Further Reading
The sources behind this post, so you can read what I read.
- Chun, Mattu, Lee and Wilkerson, “Diagnostic performance of xylazine test strips in urine from fentanyl-positive hospitalized patients,” Clinical Toxicology (2026)
- “Confusion Over Drug Tests Highlights Lack of Training for Florida Officers,” ProPublica with the Tampa Bay Times (Nov. 22, 2016)
- “FDA alerts health care professionals of risks to patients exposed to xylazine in illicit drugs,” U.S. Food and Drug Administration
This post is general information about the law, not legal advice, and it does not create an attorney-client relationship. Past results do not guarantee a similar outcome. The rule discussed is Florida Rule of Criminal Procedure 3.220(b)(1)(J) and (f). The study discussed is Chun, Mattu, Lee and Wilkerson, “Diagnostic performance of xylazine test strips in urine from fentanyl-positive hospitalized patients,” Clinical Toxicology (2026), doi:10.1080/15563650.2026.2752134. The 2014 Hillsborough figures come from Topher Sanders and Ryan Gabrielson, “Confusion Over Drug Tests Highlights Lack of Training for Florida Officers,” ProPublica, co-published with the Tampa Bay Times (Nov. 22, 2016).

