This study looks at the daylighting implications of several options for the refurbishment of an existing office building in Brussels, with a perimeter ceiling height of 2.5 m and a width of 16 m. Each side has a double-skinned facade, with a 1.4 m wide maintenance walkway in the space between the internal and external glazing. Scale models in an artificial sky, and computer simulations, were used to examine the effects of changes to the walkway. Two changes in the building＇s form were also examined, re-entrant slots in the facade, and lowering of the central area of floor. The results from these experiments are generally applicable to buildings designed with double-skinned facades, buildings using horizontal solar shading devices, light shelves or buildings with low Boor to ceiling heights. If an area of the floor space is considered to be daylit when it receives at least 300 Lux for over 50% of the working year, it was found that using the walkway options alone, the daylit area can be increased by up to 23%. Re-entrant facade slots produced no increase in the daylit area. Lowering the central floor area produced an increase of up to 14%. None of the walkway options were able to produce a daylit area of greater than 53% of the total floor space. Thus, until redirecting glazing becomes commercially viable, it is clear that shallow-plan designs are the best option for new buildings. The RADIANCE computer modelling package was used in addition to scale-model measurements. The use of both of these methods of daylighting analysis is compared. The results of both methods show good agreement.