More and more people are buying portable devices to test blood alcohol level. They're small enough to fit in your pocket, and affordable, costing as little as $50. And they're easy to use: Just plug them into your smartphone and blow to see your blood alcohol content on the spot.
The companies say these apps will help you make smart decisions. But police say they can be unreliable, and, in some cases, could lead to drunken driving.
To gauge how three such apps — Breathometer, Alcohoot and BACtrack Mobile — compare to official police Breathalyzers, the Rossen Reports team staged a social experiment: a summer barbecue and cocktail party.
Along with the guests came New Jersey State Police, armed with the official Breathalyzer they use to arrest people on the road who have blood alcohol content above the legal limit for adults of .08 percent.
Over the afternoon, one guest named LouAnn had six drinks. The police Breathalyzer showed she was drunk and would have been arrested if she had been driving: a reading of .175, more than twice the legal limit. "I can feel that I'm a definite risk," LouAnn said.
But that's not what the Breathometer said; not even close. The app said LouAnn blew .06, putting her under the legal limit for driving. Even LouAnn was surprised at the result. "Is there an app that says that's [BS]?" she asked.
Results from the Alchohoot app were wildly different: It had LouAnn at .16, twice the legal limit. "I think I agree," LouAnn said.
The third app, BACtrack Mobile, said she was even more drunk than that: .21, way over the legal limit, the highest of all three. The three apps were all over the map.
Rossen Reports: Are Breathalyzer phone apps accurate?Play Video - 5:00
Rossen Reports: Are Breathalyzer phone apps accurate?Play Video - 5:00
Rossen Reports update: Is the government fixing the tens of thousands of crumbling bridges?
Candy companies are under-filling their boxes, class-action lawsuits allege
Rossen Reports Update: What to do if you see an avalanche heading your way
Smoke alarm alert: 3 vital tips that could save your life
The party attendees were surprised. "I don't know how you trust anything if you can't trust one," said an attendee named Tali.
"I'm very surprised they're wildly different," said Jeremy, another partygoer. "It's crazy."
The police Breathalyzer found all the partygoers who had had drinks over the limit: legally drunk. The readings from all four devices:
LouAnn: Breathometer 0.06, Alcohoot 0.162, BACtrack 0.21, police 0.175
Jeremy: Breathometer 0.05, Alcohoot 0.107, BACtrack 0.143, police 0.119
Tali: Breathometer 0.07, Alcohoot 0.078, BACtrack 0.1, police 0.08
Alcohoot's numbers were closest to the police Breathalyzer's. The company called the results "impressive" and "consistent" with the police readings.
BACTrack's numbers were too high. The company told NBC News its results showcase a "small margin of error" and they "erred on the side of caution and safety for the consumer."
Breathometer's results were way too low, showing all drinkers participating in the social experiment under the limit. The company said it was "surprised" by the results, and said: "Breathometer will be releasing an all-new Breathalyzer this fall — focused on law enforcement-grade accuracy."
"Frankly, it confirms my suspicions," said Mike Halfacre, director of New Jersey's division of Alcohol Beverage Control (ABC), when he reviewed the results. "These things aren't science; they're party favors. They shouldn't be relied on. They could put people on the road when they're over the limit."
"They have to be accurate," LouAnn said. "There's too much at stake." (All the participants in the social experiment were sent home by car service.)
The companies say you shouldn't use these apps to make a decision about whether you should drive: Those warnings are on the packaging and within the apps themselves.
But police worry that that's exactly why customers will buy them: to see if they're OK to drive. They say the best advice is: If you think you need a blood alcohol test, you should already know the answer — you shouldn't drive, period.
Statement to NBC News from Breathometer:
"Thanks again for the opportunity to respond to your test results.
We are surprised by some of the results from the tests you conducted. The original Breathometer is designed to help customers and their friends with affordable insights and awareness when consuming alcohol — as well as helpful resources to make informed decisions if alcohol has been detected. The shareable form factor of the original Breathometer requires users to be familiar with the correct velocity and distance of the breath sample to obtain consistent and accurate results.
Breathometer will be releasing an all new breathalyzer this fall — focused on law enforcement-grade accuracy and enhanced features for those looking for that level of functionality. Nothing ever replaces good judgment and at no time should you drink and drive.
The Breathometer Team."
Statement to NBC News from Alcohoot:
"Alcohoot was created to help reduce the number of impaired drivers on the road by providing people with information to allow them to make smarter decisions about whether to drive when they have been drinking.
The results of the Today Show’s informal trial were very positive and the variance in the readings relatively minor. Deviation in results are common and even permitted by specifications set by the National Highway and Traffic Safety Administration even with evidential breathalyzers. Here the performance of Alcohoot is impressive, delivering results that were consistent with the New Jersey State Police results - the test subjects were over the legal limit and had they been permitted to drive would have endangered their lives and the lives of others.
Alcohoot utilizes the same fuel cell sensor and breath capturing technology that is used in law enforcement grade breathalyzers. Many other breathalyzers utilize low-grade semiconductor sensors which professionals do not endorse and warn against as they produce a wide variety of results and will often detect substances besides alcohol which produce false positives.
Utilizing this technology, Alcohoot provides accurate and actionable information to help people make considered decisions about whether or not to get behind the wheel even though it is not an evidential breathalyzer. Alcohoot can actually help prevent impaired driving since it is accessible for consumers, costs less than $100 and is simple to use. Devices used by police are not readily available to the public, cost from $800 to $5,000 and require special training to use.
Given these results, the reaction of the NJ Division of Alcoholic Beverage Control spokesperson is surprising. We expect state agencies to endorse and support tools that produce these levels of performance and enhances safety on our roadways.
Alcohoot is far from a party favor … using it might just be the biggest favor a driver can do for themselves and others on the road."
Statement to NBC News from BACtrack:
"It is great to see the TODAY Show highlight the topic of alcohol consumption.
When you consider the difference between the police results and the BACtrack Mobile results, it actually showcases a small margin of error, especially when considering the huge cost difference: $5,000+ for the police model, $129 for the BACtrack.
In no instances were the BACtrack results lower than the police model; the BACtrack results all erred on the side of caution and safety for the consumer.
BACtrack's packaging, manuals and app closely match the ABC representative's position — that users should not rely upon these results to determine their ability to safely operate a motor vehicle.
BACtrack is proud to offer an affordable way for consumers to estimate their alcohol level."