Try 1 month for 99¢

You’ve just polished off your third beer and it’s time to leave the bar. You feel a little buzzed, and you’re not sure if you should drive. So you pull out the little personal breathalyzer you bought last week at Walgreens.

You blow a .07. Good to go, right?

Maybe, maybe not.

Did you wait 20 minutes after your last drink or cigarette? Is your testing device properly calibrated for accuracy? Was it subjected to cold temperatures? Is the bar using an ion or electrical filtration air cleaner? Is your body still absorbing the alcohol in your stomach, pushing your blood-alcohol level up past the legal limit of .08 percent as you head to your car?

According to experts, these and many other factors can affect the result of your breath test, potentially rendering that sleek little device useless — or worse, giving you a false sense of security.

Nationwide, the market for personal breathalyzers is exploding. Consumers, according to Massachusetts-based WinterGreen Research, spent $215.2 million last year compared with $27.9 million in 2005, a nearly 700 percent increase in four years. And as the price drops, that could mean more breathalyzers in homes, bars and at parties.

Once almost exclusively a law enforcement tool, private consumers’ share of the breathalyzer market jumped from 7 percent to 17 percent between 2008 and 2009 and is projected to grow to 28 percent by the end of 2013, according to WinterGreen’s analysis.

In some cities you can buy a variety of models at Target or Walmart, and if they aren’t on the shelves you can get them online. For cheaper models, head to ShopKo, Walgreens or Office Max, where $40 to $50 can buy plastic, easily stowed versions.

There’s no shortage of choices. Breathalyzers come in a variety of shapes and sizes from companies like Alcohawk, Alcoscan, Lifeguard, Lifeloc and Safedrive. There’s even an iBreath that plugs into your iPhone.

Models billed as more accurate run from about $100 to $200. They use semiconductor technology, however, that still is not as precise as the fuel cell sensor models used by law enforcement in Wisconsin. The fuel cell models, which screen exclusively for alcohol, will generally set you back more than $400, but the price is dropping fast as companies scramble to cash in on the growing breathalyzer market.

Susan Eustis, WinterGreen Research president and CEO, says many of the people buying the devices are party-goers in their 20s and members of high school and college sports teams. Also, companies throwing Christmas parties often buy them in an effort to keep drunken employees off the road.

Another growing segment of the market is bars, which buy machines that allow patrons to insert a dollar, blow into a straw and see their blood-alcohol content displayed on a screen.

Eustis says sales are brisk on the East and West coasts. But 24 states have enacted stiffer drunken-driving laws since the beginning of 2009 — including Wisconsin, where new laws go into effect July 1 — and these states will likely be a “huge factor” in future sales, she says.

Police and public safety officials see it as a disturbing trend, and they shudder at the thought of thousands of drivers, fresh from bars, hopping in their cars because they blew a blood-alcohol content under the legal limit of .08. Many drivers, experts say, are impaired at .05 or below.

“It’s not the BAC level that officers look for,” says officer Deanna Reilly of the Madison Police Department, “it’s the impairment. So my concern is, if someone says ‘I’m a .07 so I’m OK to drive.’ Well, that doesn’t mean that you’re OK to drive.”

Because the devices measure alcohol, not impairment, Reilly said she thinks the practice of testing one’s self is inherently bad, even if the machines work — and she doubts that they do.

“I think they’re a waste of money,” she says.

Two weeks ago Reilly, the Police Department’s unofficial drunken-driving expert, agreed to help The Capital Times conduct an informal test to gauge two popular semiconductor personal breathalyzers against the more advanced model used by police in Madison.

Reilly showed up at the residence of a Capital Times staffer where we assembled a coterie of drinkers. Reilly is no stranger to “dosing” sessions, which she often conducts for law enforcement classes and citizen policing groups. In those situations, participants are encouraged to drink to excess so they know how it feels to be well over the limit. Our test was simply a social gathering of a bit more than three hours, where snacks were offered and participants were encouraged to drink at a pace that was more likely to put people in the neighborhood of, but not drastically above, the state’s legal blood-alcohol limit. There is, after all, little use in the real world for using a breathalyzer on someone who is obviously hammered.

Reilly came armed with an Alco-Sensor FST, manufactured by Intoximeters Inc. and available for about $500 online. She says the devices are calibrated for accuracy every two months, and the one she used was calibrated Dec. 29, about three weeks prior to the test.

The Capital Times chose an Alcomate Premium AL7000, which sells for $189.95, but which we were able to find for $142.76 on, and a Bactrack Select S30, purchased at Shopko for $39.99.

Periodically, most breathalyzers have to be sent for calibration to the manufacturer, which charges for the service. The Bactrack model purchased by The Capital Times requires calibration after 200 uses and can be serviced for a charge of $19.99. Pre-calibrated sensors can be purchased for the Alcomate, which are supposed to be installed after 200 to 300 uses, which eliminates the need to mail the device to the manufacturer.

Reilly’s device uses fuel cell technology, which is less prone to interference from substances besides breath alcohol and is less likely to register a false positive. Commercial fuel cell models boast an accuracy rate of plus or minus .005 percent. That’s twice as accurate as the two semiconductor models we purchased, which are supposedly accurate within .01 percent. They also are far more expensive, and while the price is dropping, fuel cell breathalyzers still represent a small percentage of those available to civilian consumers.

To get a clear reading, the volunteer subjects were instructed to wait 20 minutes after finishing a drink or eating any food before blowing into the three devices, as per instructions on both the personal devices. Reilly says the police device requires a 20-minute wait after consuming alcohol, but not food. The waiting period eliminates residual alcohol in the mouth as well as other substances that could result in an artificially high reading.

After the first drink the results were somewhat consistent, with the non-police versions generally registering slightly higher readings than Reilly’s device. After subsequent drinks the deviations grew, often to surprising degrees (see chart).

For instance, one participant, after drinking two Sierra Nevada Bigfoot Ales (9.6 percent alcohol by volume) registered .059 on the police device, .147 on the Alcomate and .11 on the Bactrack (the Bactrack displays only two decimal digits, while the others display three). The third beer took the drinker out of the driver’s seat on all three devices, but again, by vastly different margins: .091, .113 and .14 respectively.

Another participant, after drinking three New Glarus Brewery Spotted Cows (at 5.1 alcohol by volume, roughly the same as Budweiser and slightly higher than Miller Genuine Draft) blew .049 on the police breathalyzer, .071 on the Alcomate and .09 on the Bactrack. After four, he registered .074, .085 and .13, respectively.

Mort Parto, sales manager for Alcomate, says the test was invalid because it was conducted on people, not scientific equipment. And he questions whether the police model was properly calibrated.

“We really don’t know who’s accurate, ours or the others,” he says.

(Just for the record, there’s no scientific way for us to check which device is the most accurate. But The Capital Times found that the blood-alcohol calculator on the Wisconsin Department of Transportation’s website, which gives estimates based on weight, gender, number of drinks and drinking time frame, was in most cases uncannily consistent with the police readings.)

Parto also questions the practice of testing a fuel cell device against semiconductor models.

“You’re comparing two units that are a different class of technology,” he says.

Asked what confidence consumers can have in the semiconductor machines he sells if they don’t compare with more accurate models, he notes that none of the hand-held devices is used to record blood-alcohol levels that are admissible in court. “This is just for safeguarding and screening,” he says.

Sometimes, they’re just for amusement.


“To be honest with you, it’s just a game,” says a man who answered the phone as the owner of Buck’s Madison Square Garden near the UW campus, but then identified himself when asked for his name as “Royal Doolittle.” (The actual owner of Buck’s is James Dailey.)

The bar is a popular spot for college students and young adults, who often slip a dollar into a breath test machine to see who scores the highest, he says, without bothering to read the instructions, which say to wait 20 minutes after drinking to get an accurate reading.

“To me it’s nothing but a vending machine,” the man says. “It might as well be a dart board.”

While the results of The Capital Times’ in-home test call into question the accuracy of consumer-grade, semiconductor breathalyzers, local attorneys who represent accused drunken drivers say there is plenty of reason to question the ones used by police as well.

When used in the field, police breathalyzer tests are legally referred to as preliminary breath tests, or PBTs, and they are not admissible as evidence in court. In Madison, the machine officers use to get evidence that will hold up in court resembles a typewriter and sits on a desk in the Police Department. It uses a combination of fuel cell and infrared technology, and those under suspicion of drunken driving take two tests in three minutes. Those tests have to come within 0.002 percent of one another. In Dane County, those suspected of a third offense or higher, or suspected drunk drivers who have been involved in an accident involving death or injury, are generally required to have their blood drawn, which is the gold standard of blood-alcohol testing.

As Madison as it gets: Get Cap Times' highlights sent daily to your inbox

In 2007, Madison attorney Tracy Wood managed to get a conviction for a third-time drunken-driving offense overturned — even though a forced blood test showed a blood-alcohol concentration of .191 — in part because PBTs are not deemed sufficiently accurate.

The state Court of Appeals decision revolved around the suspect’s refusal to take a PBT. Walworth County prosecutors used his refusal to win a jury conviction, but the appeals court ruled that if you can’t use a PBT in court, you can’t use the refusal to take a PBT in court, either.

“The concern here is that a jury is generally unaware of the scientific imperfection of the PBT device, and that it might well be more influenced by proof of a driver’s refusal to take the test than by the evidence of adverse results coupled with proof of its scientific imperfection or inaccuracy,” the appeals court wrote.

The reading from the PBT is basically useless, Wood says.

“I think what they are good for is telling whether or not someone’s had a drink — if they’ve had something to drink with ethanol in it. That’s about all the reliability I would say they have.”

Officers are under strict guidelines when using the devices. In 1999, local defense attorney Stephen Mays took a case to the state Supreme Court that led to the current standards for police procedure. The court said officers have to conduct field sobriety tests such as looking for involuntary jerking of the eyes, walks and turns and one-leg stands, then, if the officer has a “reason to believe” a driver is intoxicated, the officer can administer a preliminary breath test to help determine whether to make an arrest. If an attorney can show that the test was administered improperly, it could provide a means to get a conviction overturned.

“If you’ve got a guy who fails the eye test, fails the walk and turn, fails the one-legged stand, you really don’t even need this device,” Mays says. “It’s where it’s questionable. The guy passes two out of three, then you suspect. Then you want to use it. But then the big question is do you have a reasonable belief at that time if they pass two out of three? So we do litigate these things a lot.”

If you can’t trust the breath tests used by police, why should you trust the ones you buy at Walmart or on the Internet?

“These things are convenient and affordable, but they are far more likely to provide inaccurate and inconsistent readings than the device used by law enforcement,” says Beth Mosher, a spokeswoman for the American Automobile Association.

Even if they worked, she says, they’re a bad idea.

People are “using them when they already have alcohol in their system, so their judgment in some cases is already impaired,” she says.

Sue Hackworthy couldn’t agree more.

Hackworthy, chief of the chemical test section of the state Department of Transportation, is in charge of testing and approving all breath-alcohol instruments used by law enforcement agencies in the state.

She says someone in the highway safety division bought one of the consumer-grade semiconductor breath analyzers and suggested that “it would be kind of neat” for DOT personnel to hand the devices out to people.

She took it to her lab and conducted a brief test.

“It was all over the board,” she says.

She didn’t bother to do more extensive testing because she already had decided “this is a bad idea.”

For one thing, she has little confidence that people will use them properly or keep them properly calibrated. Hackworthy also notes that everyone absorbs and eliminates alcohol at different rates, and those rates vary depending on how much food they’ve eaten, how quickly they are consuming alcohol and a variety of other factors. It takes 30 to 90 minutes to fully absorb alcohol, meaning if you drink alcohol at a faster rate than your body eliminates it, you can keep getting drunker and drunker after you’ve finished your last drink.

And by that time, you might be well down the road.

Sign up to receive free subscriptions to our email newsletters

Never miss anything! Stay updated with our latest news, commentary and other features via e-mail.

Your information will not be shared with third parties.

Please see our Privacy Policy and Terms of Service for more information.

I understand and agree that registration on or use of this site consitutes agreement to its user agreement and privacy policy.

Subscribe to Breaking News

* I understand and agree that registration on or use of this site constitutes agreement to its user agreement and privacy policy.