In this paper we describe a method for the preliminary assessment of the illumination needs of partially sighted individuals. The method is based on determining the smallest readable lettersize under different levels of illumination. Half of the partially sighted in our population had a performance that was dependent upon the level of illumination. At the same time, most of those for whom no improvement in performance was found, did express a preference for a certain level of illumination. Therefore, there remain discrepancies between the objectively determined optimal level and the preference for a certain level. We conclude that the current method is a useful but insufficient method of determining the illumination needs of partially sighted individuals.