Subjects were asked to take part in experiments which involved simulated escapes through a network of corridors. The purpose was to determine a safe minimum illuminance which should be provided by escape lighting systems. A substantially uniform floor illuminance of 0. 28 lux was found to give subjects adequate visibility for carrying out the experimental task. The effects of a diversity in the emergency illuminance and disability glare were also noted. Within the limits investigated these were not found to affect performance. It was considered that a form of glare limit should be proposed to ensure the legibility of escape signs under emergency conditions.
