PURPOSE: To determine how simulated visual impairment impacts on night-time driving performance and pedestrian recognition measured under real road conditions. EXPERIMENTAL DESIGN: Night-time driving performance of 20 young normal participants (M=27.5 � 6.1 yrs) was assessed on a closed road circuit for three different visual conditions. The visual conditions were mounted in modified goggles and contained simulated cataracts, refractive blur selected to match the visual acuity of the cataract condition and normal vision. All participants had binocular visual acuity greater than 6/12 (20/40) when wearing the goggles and satisfied the visual requirements for driving. Driving measures included road sign recognition, detection and avoidance of low contrast road hazards, lane-keeping and time to complete the course. Participants were also scored on recognition of pedestrians who wore either black clothing or retroreflective markings on either the torso or the limb-joints to create "biomotion." RESULTS: Simulated visual impairment reduced night-time driving performance (p[[lt]]0.05); posthoc testing indicated that these differences were greatest for the cataract condition, even though the cataract and blur conditions were matched for visual acuity. While visual impairment significantly reduced the ability to recognise the darkly clad pedestrians, the pedestrians wearing "biomotion" were seen 80% of the time. CONCLUSIONS: These data confirm that driving performance under night-time conditions is significantly degraded by the effects of early visual impairment and that pedestrian recognition is greatly enhanced by marking limb-joints to create "biomotion" which was relatively resistant to the effects of visual impairment.