1.) The assumed "partition ratio" of 2100, which converts the breathalyzer reading to a predicted BAC, actually varies between 1300 and 3100 according to Wikipedia. This ratio is based on Henry's Law and the van 't Hoff equation. Based on the Henry's Law Constant for Ethanol and the partition ratio of 2100 assumed for lung air temperature of 34C, the constant for the Van Hoft Equation can be determined. Applying this to other temperatures shows the exponential dependence of the partition ratio on temperature and explains why the actual partition ratio varies so widely in people.
(The content of this paragraph has been removed. We encourage readers to check out the section on common sources of error at Wikipedia for additional information.)
2.) Hyper-ventilating prior to sampling has also been shown to lower breathalzyer apparent BAC by 22%. This is most likely no different than number two (2). This is because air temperature is normally less than body temperature - hyper-ventilating increases air flow, cooling the lungs more than normal resulting in lower lung air temperatures and a lower partition ratio. Don't try it if it's hotter than 98F.
3.) Activated charcoal is one other theory (more commonly purported than the one we advance above) that hasn't been completely refuted by scientists, Wikipedia, and MythBusters.
1.) Mouth alcohol is the easiest way to induce a false positive breathalyzer reading. While mouth alcohol generally dissipates in 7 - 10 minutes, most breathalyzer manufacturers and law enforcement agencies recommend waiting 20 minutes from the last drink before sampling. Dentures and advanced gum disease can trap mouth alcohol and extend the time needed for mouth alcohol to clear.
2.) Gastro-intestinal reflux / hiatal hernia can allow naturally occurring alcohols in the stomach to enter the mouth and mix with the breath. Burping can also allow stomach alcohols into the breath to a lesser extent.
3.) Interfering compounds, such as acetone, can be misinterpreted as alcohol by some breathalyzers. This is especially problematic for diabetics.
4.) Some "remedies" people use to "fool" breathalyzers can cause false positive / artificially high breathalyzer readings. Mouthwash can be particularly damaging since it contains alcohol.
5.) Copper and silver in the mouth are nice in theory since both can be catalysts for converting alcohol to other products but in practice neither are effective for converting alcohol in the breath on its way out of the mouth.
6.) Running a high fever or otherwise raising your "partition coefficient" can also cause a false positive. More on this on the "false negative" section.