Mock-up model can be applied to measure accurate performance data but difficult to apply the variables in experiment. There can be a slight experiment errors in Scale model, but various parameters can be applied for a objective experiment. This paper aims to compare the daylighting performance in 1, 1/5, 1/10 scale model of offices and analyze the experiment errors to certificate the influence of model experiment. To analyse daylighting performance, a comparison of a Mock-up model, sized 12.0m(w)x7.2m(l)x3.7m(h), designed for experimentation of day lighting systems and its 1:5, 1:10 scale model. Measurement was conducted at a point viewing the window part from the rear end of outdoor center-line so that window part, ceiling plane, left and right wall plane and floor plane could be measured. And luminance of window, rare of the room was measured under clear sky. It is to be monitored by the Radiant Imaging ProMetric 1400.