Breath-test evidence center

The number is the start of the review.

A Florida-first guide to the machines, safeguards, maintenance records, expert analysis, misconceptions, and source-code cases behind evidential breath testing.

Florida rules checked September 1, 2026 Official-source links Other states clearly labeled

Educational information—not legal or scientific advice. Do not use this page to decide whether to take or refuse a test, miss a hearing deadline, or represent yourself. Refusal and license rules can carry immediate consequences. The exact instrument, software, forms, law, and evidence must be checked for the test date and jurisdiction with a licensed attorney and, when appropriate, a qualified expert.

01 / The foundation

Roadside screening is not the same as an evidential test.

People often call every breath device a “Breathalyzer.” In practice, a handheld preliminary breath tester and an evidential instrument at a station serve different roles, operate under different programs, and may be treated differently in court.

Preliminary breath tester (PBT)

A portable roadside device is commonly used as an investigative screening tool. Many use a fuel-cell sensor. Its use, admissibility, calibration schedule, and refusal consequences depend on state law and the purpose for which the result is offered.

Evidential breath tester (EBT)

A larger approved instrument runs a prescribed test sequence and creates records intended for evidentiary use. Florida's approved method uses infrared light absorption, and its current approved list names the Intoxilyzer 8000 and Intoxilyzer 9000 with Department-evaluated software.

What the instrument actually reports

Breath instruments analyze alcohol in an exhaled breath sample and report a breath-alcohol concentration—typically grams of alcohol per 210 liters of breath. That is not a direct laboratory measurement of blood. The law defines the breath threshold and supplies rules for how a qualifying result may be used. Florida also recognizes an independent test that the tested person may arrange at personal expense after the required test, subject to the statute's conditions.

02 / Machine deep dive

Every device must be matched to its exact configuration.

A model name alone is not enough. The meaningful review includes the sensor method, approved software or firmware, serial number, jurisdiction, required test sequence, operator permit, inspection history, standards used, and the complete electronic record.

Florida-approved

CMI Intoxilyzer 8000

An infrared evidential instrument with a long Florida deployment history and an extensive FDLE records archive.

Program strengths
Required air blanks, diagnostics, control testing, multiple subject samples, recurring inspections, and public instrument records.
What deserves review
The precise sample sequence, agreement between results, operator actions, error or abort messages, repair history, standards, and software version.
Florida-approved

CMI Intoxilyzer 9000

Florida's newer approved infrared platform, now covered by current operational and inspection forms.

Program strengths
The current sequence includes a 20-minute observation period, blanks, diagnostics, control checks, and an additional sample when the first two do not agree within the rule's tolerance.
What deserves review
Correct model-specific forms, operator training, software approval, transition records, controls, result agreement, and instrument history.
Used in other states

DataMaster DMT / Intox DMT

An infrared instrument family; some jurisdiction-specific configurations also use a fuel cell.

Potential strengths
Automated sequencing, diagnostics, internal records, and—where configured—two distinct sensing approaches.
What deserves review
The exact DMT configuration, software, state approval, simulator or dry-gas controls, observation rules, calibration history, and data files.
Used in other states

Intox EC/IR II

Uses an electrochemical fuel cell to quantify alcohol while infrared measurements monitor alcohol, carbon dioxide, and the breath profile.

Potential strengths
Independent sensing information, breath-profile monitoring, automated checks, and tools intended to identify mouth alcohol or inadequate samples.
What deserves review
The state's approved configuration, sensor checks, consumables, records, environmental conditions, operator procedure, and flagged breath profiles.
Used in other states

Dräger Alcotest 9510

An evidential platform that combines infrared measurement with an optional electrochemical sensor, depending on configuration.

Potential strengths
Two-sensor comparison where enabled, automated diagnostics, interferent checks, and configurable evidential workflows.
What deserves review
Approved firmware, local protocol, calibration and certification records, sensor agreement, status messages, and the complete test record.
Roadside screening

Portable fuel-cell devices

Handheld units estimate alcohol from a small breath sample and can be useful for rapid screening in the field.

Potential strengths
Portable, fast, comparatively specific fuel-cell response, and useful for investigative screening when properly maintained.
What deserves review
Whether the device was approved for the claimed use, calibration interval, operator training, sampling conditions, and the limited purpose allowed by local law.

This is a representative comparison—not a nationwide approved-products list and not a finding that any listed model is inherently accurate or inaccurate. NHTSA publishes model specifications and conforming-products lists; each jurisdiction decides which instruments and configurations it approves.

03 / Proper protocol

A reliable review follows the entire test—not one printout.

Florida's operational forms are model-specific. The outline below captures the core current safeguards, but the actual approved form and rule in effect on the test date control. Florida applies a substantial-compliance standard, so a deviation does not automatically equal exclusion.

Qualified operator and approved instrument

Confirm the operator's permit, instrument make and model, serial number, registration, approved software, and whether the instrument was in evidentiary service.

Observation before sampling

Florida's current operational procedures use a minimum 20-minute observation period. The evidence should show when it began and what occurred during it.

Blanks, diagnostics, and controls

The sequence checks for a clean air path, successful instrument diagnostics, and an acceptable result from a known alcohol standard before relying on a subject result.

Two valid subject samples

The instrument seeks separate samples. Under the current 9000 procedure, a third is requested automatically when the first two do not agree within 0.020 g/210L.

Post-test safeguards

Model-specific procedures run additional control, blank, or diagnostic checks after the subject samples. Review their values and any status messages.

Preserved result and affidavit

Compare the printout or affidavit with electronic data, timestamps, video, operator testimony, inspection files, repair entries, and nearby tests on the same instrument.

04 / Maintenance and records

The instrument history can matter as much as the test day.

Florida currently requires agency inspections at least once each calendar month and Department inspection at least once every 12 months. A Department inspection is also required after repair and before the instrument returns to evidentiary use. Current rules impose multi-year retention requirements for key inspection, repair, registration, and standard records.

Document packet to discuss with counsel

  • Subject-test affidavit, printout, and electronic data
  • Video and timing evidence for the observation period
  • Operator permit, training, and model authorization
  • Instrument registration and approved software version
  • Agency inspection reports before and after the test
  • Latest Department inspection and registration packet
  • Repair, maintenance, removal, and return-to-service records
  • Dry-gas or alcohol-solution certificates and lot information
  • Diagnostics, controls, blanks, error codes, and aborted tests
  • Nearby subject and inspection tests that provide context
  • Current operational form, inspection forms, and manuals
  • Chain-of-custody or handling records when relevant

FDLE publishes Intoxilyzer 8000 registrations, Department inspection packets, agency inspections, repair materials, correspondence, and electronic test data online. A public file is a strong starting point, but counsel should determine what additional discovery is available and relevant.

05 / Expert witnesses

An expert tests a theory against records and science.

The responsible question is not “Can an expert create doubt?” It is “Can a qualified expert identify a scientifically supportable limitation, inconsistency, or compliance issue that matters under the law and facts?”

What a qualified expert may help evaluate

  • Whether the approved test sequence and quality controls were completed and documented.
  • Whether two results agree, and what any third sample, error, abort, or status message means.
  • Whether inspection, repair, standard, or return-to-service records reveal a relevant issue.
  • Whether the operator's testimony matches the electronic record, video, and timestamps.
  • Whether a medically documented physiological factor is scientifically relevant to the breath result.
  • Whether a software, firmware, calculation, or data-integrity question is specific enough to justify further review.
  • How uncertainty, measurement limitations, and the timing of drinking relate to the opinions actually offered.

What an expert cannot promise

  • A guaranteed dismissal or exclusion.
  • That every missing document invalidates a test.
  • That every breath machine is unreliable.
  • That a source-code request is always available or necessary.
  • An opinion that outruns the records, accepted methodology, or the expert's qualifications.
  • A substitute for legal advice from defense counsel.
06 / Common misconceptions

Neither “perfect” nor “junk” is a useful starting point.

Breath evidence should be evaluated in the disciplined middle: identify what was measured, confirm the approved process, inspect the complete record, and connect any claimed problem to the result and governing law.

Myth 01“A breath test is the same as a blood test.”

Reality: Breath and blood tests measure different specimens and report different concentration units. The legal threshold for breath is defined in grams per 210 liters of breath.

Myth 02“Two results have to be identical.”

Reality: Programs use an agreement tolerance, not perfect identity. The current Florida 9000 procedure calls for a third sample when the first two are not within 0.020 g/210L.

Myth 03“Any missed step automatically dismisses the case.”

Reality: Florida law uses substantial performance of approved methods and says insubstantial differences do not automatically invalidate a result. Materiality and remedy are legal questions.

Myth 04“A machine number can never be questioned.”

Reality: Approval does not prevent review of the operator, sequence, instrument records, standards, samples, maintenance, or the meaning and weight of the result.

Myth 05“No source code means the result is automatically invalid.”

Reality: Discovery and admissibility depend on the jurisdiction and a specific showing. Florida's Ulloa decision addressed how to subpoena out-of-state materials; it did not decide scientific accuracy.

Myth 06“An expert is always—or never—worth hiring.”

Reality: Value depends on the stakes, reading, records, disputed facts, potential issue, expert qualifications, and defense strategy. Counsel should define the question before retaining an expert.

07 / Court decisions

Read the holding—not the headline.

Breath-test decisions are device-, software-, record-, date-, and jurisdiction-specific. A case may concern discovery, scientific reliability, calibration, or remedy; those are not interchangeable questions.

Florida Supreme Court · 2013

Ulloa v. CMI, Inc.

Defendants sought Intoxilyzer 8000 source code held by the out-of-state manufacturer. The court held that Florida's Uniform Law procedures had to be used to obtain out-of-state, nonparty materials. It expressly did not decide whether the source code would be admissible or whether the machine was scientifically accurate.

New Jersey Supreme Court · 2008

State v. Chun

After extensive expert and special-master review, the court found New Jersey's configured Alcotest 7110 scientifically reliable subject to procedures and safeguards. The later State v. Zingis opinion recounts that framework and its continuing record requirements.

New Jersey Supreme Court · 2018

State v. Cassidy

The court held that results from instruments calibrated by an officer who failed to use a NIST-traceable thermometer were not sufficiently reliable and were inadmissible. The decision illustrates why actual calibration records and required safeguards can matter.

The careful takeaway

Source code is one possible issue—not the whole review.

Courts may require a case-specific basis, proper discovery procedure, protective conditions, or proof of materiality. Often the more immediate evidence is the test sequence, electronic data, controls, inspections, repairs, approved software, operator actions, and video.

Important distinction: A source-code dispute does not itself prove a wrong result, and the absence of source code does not automatically make a test admissible. Counsel must connect the requested material to a live factual or legal issue and use the discovery procedure available in that jurisdiction.
08 / Primary sources

Start with the current official record.

Rules and forms change. These sources were checked on September 1, 2026. Always confirm the version in effect on the date of the test.

Bring specific questions—not assumptions—to counsel.

Use this guide to identify the instrument, collect the right records, and understand the questions a lawyer or qualified expert may need to investigate.

Read the jury-trial guide →