The legal limits of blood alcohol concentration (BAC) for operators of commercial vehicles throughout the United States, and underage drivers in most states, are approximately half the 0.08–0.10% imposed on other drivers. Easily administered measures involving performance and appearance are needed to establish probable cause for requesting breath tests. Several measures showing relationships to blood alcohol were examined for their ability to distinguish BACs under and over 0.04% within the 0.00–0.08% range. Measures of heat loss, pupillary response, balance, complex tasks, and accuracy under speed were studied in controlled experiments with alcohol-dosed subjects. The only reliable index of blood alcohol was horizontal gaze nystagmus (HGN) as administered in the standardized field sobriety test (SFST), using alternative scoring criteria. Although other measures evidenced intra-individual change at low BACs, the magnitude of change was greatly exceeded by inter-individual differences, thwarting their use in detecting drivers with BACs lower than 0.08%. HGN, however, proved as valid in detecting BACs in the 0.04–0.08% range as at the higher levels of the SFST. It is also as valid when administered to a seated subject as one standing, making it of particular benefit in enforcement, where the behavior of low BAC drivers rarely provides probable cause to request that they leave a vehicle to be tested in a standing position.