Abstract
A real road experiment has been conducted in order to investigate how to represent landmarks for guidance systems. The question is to define design criteria for these landmarks in order to be understood and usable by drivers. In this framework, two types of pictorial designs have been tested: generic type of message or specific message. Objective data such as drivers' visual strategies and subjective data in terms of preference have been analysed with the aim to give some recommendations for the future design of these landmarks. (A)