This study investigated real-world use of high beams using a secondary analysis of data from a field operational test (FOT) of two crash warning systems. This FOT featured 11 instrumented vehicles and 87 participants. With the assistance of the Michigan Secretary of State office, 6,000 licensed drivers were selected at random for possible participation in the FOT. The random sample contained only drivers who belonged to one of the following age groups: 20-30, 40-50, or 60-70 years. The drivers were instructed to use instrumented experimental vehicles as their personal vehicles for extended time periods. The data collected during night time driving provided information about high-beam use on a wide variety of road types. Drivers in this FTO study participated of two crash warning systems and were unaware that data about their high-beam use were being collected drivers were given instrumented vehicles to drive as their own car for periods ranging from 7 to 27 days. Most of the driving was in south-eastern Michigan. The results indicate that high-beam headlamp use is low, a finding consistent with previous studies that used different methods. Combined over all drivers and all road types, high beams were used for 3.1% of the distance driven at night. Drivers used high beams most frequently on local roads, but even that use was only 8.9%. Under the most favourable conditions for high-beam use (rural roads with, no opposing or preceding vehicles), drivers used their high beams 25.4% of the night time distance driven. Older drivers used high beams three times more frequently than did younger drivers, suggesting that older drivers, at least to some extent, understand their visual impairment at night, and therefore take some compensatory measures. Because low beams do not provide sufficient illumination in many driving situations, the present findings imply that (1) increased high-beam use should be encouraged and (2) the use of automatic switching between high and low beams is likely to be beneficial.