The theory that blood (containing alcohol) present in the oral cavity may falsely increase breath analysis recently led to a successful appeal against a drink driving conviction. Subjects who had previously consumed vodka (37.2% alc/vol), at 30 ml/10 kg and reached a BAC (blood alcohol concentration) of between 0.05 and 0.10% were then given four oral solutions consisting of a control (distilled water), and 0.05, 0.10 and 0.15% aqueous alcohol (ethanol) solutions, administered in coded form. A four-way cross-over, blind, randomized assay was conducted with the solutions, with breath analyses conducted in the presence or absence of solution in the mouth. The first trial group (n = 18) received 2 ml of solution, and we found that the simulated 0.15, 0.10 and 0.05% alcohol solutions in the mouth produced BAC reading increases of 0.0088 +/- 0.0014, 0.0062 +/- 0.0008 and 0.0055 +/- 0.0010% respectively (p < 0.001). The second trial group (n = 20) received 1 ml of solution and produced BAC reading increases of 0.0047 +/- 0.0011 (p < 0.001), 0.0023 +/- 0.0008 (p < 0.01) and 0.0020 +/- 0.0006% (p < 0.05) respectively. In conclusion, these studies indicate that small volumes of blood (containing alcohol) in the mouth would not have a practical effect on breath analysis readings.