Office lighting has been suggested as one of the possible factors in producing 'building sickness'. Health questionnaires were completed by 106 out of 109 (97%) workers in six randomly sampled multi-occupied offices in each of two buildings, one air-conditioned and one naturally ventilated. There was a significantly higher prevalence of work-related headache and work-related lethargy in the air-conditioned building than in the naturally ventilated one. There was also less daylight in the air-conditioned building and lower mean luminance and illuminance of the work positions despite there being more lights on (p<0.01). The workers had a greater dislike of fluorescent lighting (p<0.01) and overall found the lighting to be less comfortable (p<0.01) and glare readings were higher. The workers perceived their control of lighting as poorer (p<0.001) and consequently there was less agreement about it (p<0.001). Those with work-related headache found the lighting less comfortable (p = 0.059) and perceived more glare (p<0.05). This study suggests the need to maximize the use of natural light from untinted windows, to reduce the impingement of fluorescent tubes on the line of sight and to return the control of levels of lighting to each individual worker.