A breath test result feels objective, like a number from a scientific instrument. But that instrument is only reliable if it is calibrated and maintained exactly as the rules require, and very often it is not. I am Joel Brand, and attacking the calibration and maintenance records is one of the most productive things I do on a DUI.
A breath machine is only as good as its upkeep
A breathalyzer does not measure alcohol directly. It infers your blood alcohol from a sample of your breath using assumptions and a calibrated reference. If the machine has drifted, if its calibrating solution or gas standard is off, or if it has not been checked on schedule, the number it produces is not trustworthy. Jurors tend to assume the machine is infallible. It is not. It is a piece of electronic equipment that needs regular, documented servicing to produce a result anyone should rely on, and that servicing is governed by detailed regulations precisely because the stakes are so high.
What the rules require
California's Title 17 regulations require breath-testing instruments to be calibrated and accuracy-checked on a strict schedule, with records kept for each device. Under those rules, an evidentiary breath instrument generally must be accuracy-checked every 10 days or every 150 blows, whichever comes first, and the results have to stay within a defined tolerance of the true value. The agency has to follow its own maintenance protocol and keep complete logs. When it does not, the reliability of your result is fair game, and the prosecution's central piece of evidence starts to wobble.
The 15-minute observation and procedure
Title 17 also governs how the test is given, not just how the machine is maintained. The operator is supposed to continuously observe you for about 15 minutes before the test to ensure you do not burp, regurgitate, eat, drink, or smoke, any of which can introduce mouth alcohol and inflate the reading. Two separate breath samples are required, and they must agree within a set tolerance. When the observation period is skipped or cut short, or the two samples disagree, the result is open to challenge. This is where the calibration attack overlaps with the mouth alcohol defense.
Where these cases break down
Missing logs, late accuracy checks, devices that drifted out of tolerance, repair records that show recurring problems, and incomplete maintenance histories all undermine the number. I subpoena and comb through those records, because a result from an improperly maintained machine should not be treated as gospel. I look for accuracy checks that were overdue, readings that fell outside tolerance and were not addressed, service visits that suggest the unit was malfunctioning, and gaps where required records simply do not exist. Any one of these can be the thread that unravels the breath result.
The right to preserved evidence
Defendants also have a right to preserved evidence, which gives real leverage when records are incomplete. If the agency failed to retain the calibration logs, the maintenance records, or, in a blood case, the sample itself, that failure can be used to argue the result is unreliable and, in some situations, to exclude it. The burden is on the state to show its evidence was produced reliably. When the paperwork is missing, that burden becomes very hard to meet.
Blood tests are not immune either
People assume a blood draw is bulletproof, but blood evidence has its own failure points. The sample must be drawn by a qualified person, mixed with the correct preservative and anticoagulant, stored at the right temperature, and analyzed properly. Improper storage can allow fermentation, which actually generates alcohol in the vial and pushes the reading up. Clerical and chain-of-custody errors happen too. The same demand for complete, accurate records that drives the breath calibration attack applies with full force to blood results.
The partition ratio assumption
There is a deeper scientific weakness built into every breath test. To convert the alcohol in your breath into an estimate of the alcohol in your blood, the machine relies on a fixed assumption called the partition ratio, typically set at 2,100 to 1. The problem is that this ratio is an average, and it varies significantly from person to person and even within the same person depending on temperature, breathing pattern, and physiology. Someone whose true ratio is lower than the assumed average will register a falsely high BAC. A breath result is therefore not a measurement of your blood at all; it is an estimate built on an assumption that may not fit you, which is one more reason these numbers deserve scrutiny rather than blind trust.
How I get the records and use an expert
The calibration and maintenance logs are not handed over automatically. I serve targeted discovery and subpoenas to obtain the full service history, the accuracy-check records, the usage logs, and the operator's training and certification. Where the records reveal a problem, I bring in a forensic toxicologist or breath-testing expert who can explain to the prosecutor, the judge, or a jury exactly why the number cannot be trusted. Prosecutors take these challenges seriously, because once the reliability of the chemical evidence is genuinely in question, the foundation of the whole case is shaken, and that often translates into a dismissal or a favorable reduction.
How it fits with the rest of the case
The calibration attack pairs naturally with the mouth alcohol and rising BAC defenses, and certain medical conditions can compound an already shaky reading. It is just as relevant at the DMV hearing as in court, because the DMV also relies on that breath or blood number to justify the suspension. It applies squarely to a VC 23152(b) charge, where the number is the entire case. See my top DUI defenses.
Think this applies to you? Let's find out.
Whether this defense fits depends on the specific records and facts of your stop, which is exactly what I review. Use the free case analysis on this page, or call me directly at (888) 271-6644. I answer my own phone, 24/7.
From the DUI blog: What to Do When the Police Got Your BAC Wrong at the Scene.