Abstract
In this paper a simple "on road" test procedure is described which enables the performance of bicycle lighting systems to be assessed quantitatively. The results suggest that the most desirable beam pattern incorporates an intense spot on the road, far ahead of the bike, with a broad light pool filling the area between the spot and the front of the bike. While none of the commercially available systems which were tested could match this performance, the results demonstrated that the development of significantly improved headlight designs is well within the scope of currently available technology.