Anecdotal reports suggest that high environmental or occupational exposures to the fuel oxygenate methyl tert-butyl ether (MTBE) may result in breath concentrations that are sufficiently elevated to cause a false positive on commercial breath-alcohol analyzers. We evaluated this possibility in vitro by establishing a response curve for simulated breath containing MTBE in ethanol. Two types of breath-alcohol analyzers were evaluated. One analyzer's principle of operation involves in situ wet chemistry (oxidation of ethanol in a potassium dichromate solution) and absorption of visible light. The second instrument uses a combination of infrared absorption and an electrochemical sensor. Both types of instruments are currently used, although the former method represents older technology while the latter method represents newer technology.The percent blood alcohol response curve was evaluated over a breath concentration range thought to be relevant to high-level environmental or occupational exposure (0-361 microg/l). Results indicate that MTBE positively biases the response of the older technology Breathalyzer when evaluated as a single constituent or in combination with ethanol. We conclude that a false positive is possible on this instrument if the MTBE exposure is very high, recent with respect to testing, and occurs in combination with ethanol consumption. The interference can be identified on the older technology instrument by a time dependent post-reading increase in the instrument response that does not occur for ethanol alone. In contrast, the newer technology instrument using infrared and electrochemical detectors did not respond to MTBE at lower levels (0-36 microg/l), and at higher levels (>72 microg/l) the instrument indicated an "interference" or "error". For this instrument, a false positive does not occur even at high MTBE levels in the presence of ethanol.