A variety of breads and soft drinks were tested and found to contain low concentrations of alcohol. The potential for these products to generate false readings on an evidential breath-alcohol instrument was evaluated. Alcohol-free subjects ingested these products and then provided breath samples into a DataMaster. It was found that breath samples provided immediately after consumption of some of these products, or with them still present in the mouth, did produce low levels of apparent breath alcohol, which may or may not be rejected as invalid by the breath-test instrument. If the subject swallowed or expectorated the food or beverage and then observed a 15-min deprivation period during which nothing was introduced into the mouth, the apparent effect was eliminated. These findings emphasize the need for the mandatory pretest alcohol-deprivation period and the benefits of duplicate breath sampling.