From a link at Insty ran across several interesting things

at a blog:
Eating bread can cause a breathalyzer to indicate you've been drinking
After consuming almost any type of bread product — white loaf bread, donuts, pretzels, pastries, etc. — Price consistently registered blood-alcohol readings on the machine. These levels were commonly around .03%, but rose as high as .05% (enough, along with a drink or two, to reach illegal levels). Further, the Intoxilyzer’s slope detector (an electrical circuit designed to detect alcohol from the mouth rather than from the lungs) failed to indicate the presence of any "mouth alcohol". He reported this in an article entitled "Intoxilyzer: A Bread Testing Device?", 15(4)Drinking/Driving Law Letter 52.

Reacting to the use of this article by defense attorneys in their state, the Washington State Toxicology Laboratory conducted their own studies to refute the findings — this time with the machine used in Washington, a DataMaster. Unfortunately, their research only confimed Price’s experience.
More details, well worth reading. As is this:
I’ve written in the past about how most so-called "breathalyzers" do not measure alcohol: they actually measure the presence of a molecular group in chemical compounds. Ethyl alcohol (aka ethanol) contains the group, and so when the machine detects its presence (or, more accurately, infrared energy is absorbed by it), it simply assumes that the detected compound must be ethyl alcohol.

Problem: there are thousands of compounds containing the molecular group — of which well over one hundred have been found on the human breath.

Breathing gasoline or paint fumes, for example, or merely absorbing the fumes through the skin, can create false breath test results for days afterwards. And I’ve posted in the past that the problem is particularly acute when the suspect happens to be a diabetic, as diabetics often have high levels of acetone in their breath — a compound which contains the group in its molecular structure.

And, just to demonstrate once again what a nanny-state bastard Sen. Schumer is,
U.S. Sen. Charles Schumer, D-N.Y., was in Elmira on Friday to promote a bill that would fund research of alcohol detection devices that could make such a scenario possible. Called the ROADS SAFE Act of 2010, or Senate bill 3039, the act would direct the National Highway Transportation Safety Administration to spend $60 million over five years to figure out whether it’s possible to make an alcohol-detecting ignition lock that is both reliable and affordable.
Besides the general problems with this(are the going to make it illegal to start the car if you're wearing gloves? Etc.), I have to wonder if those compounds mentioned above are also found in sweat? Which would cause false readings and make someone unable to start the car(but that's not a problem because it's All For Your Own Good, no matter how much trouble it causes)...