Optital properties,of window roller shades, which affect daylight performance, visual comfort, and energy use, are usually assumed constant or diffuse. This paper presents a comparison of modeling approaches for calculating shade optical properties and the potential effects on daylighting performance and visual comfort. An angular empirical model, only requiring basic inputs provided by manufacturers (openness factor and total normal visible transmittance), showed good agreement with measured data (integrated sphere measurements and measured BTDF data)with modifications needed only for high incidence angles. The impact of fabric modeling approaches on daylight distributions and glare evaluation was studied using four different shade modeling methods, comparing with full-scale experimental data and with annual daylight simulation results. The results confirm that predicting correct shade optical properties is an important component of daylighting calculation modules. Current "constant transmittance" or "non-angular" models may result insignificant errors in daylight metrics and glare probability. Finally, the basic properties of 14 roller shade products were measured and compared with manufacturer's data. Most listed data are inaccurate and there is no standard for accurate product specifications. It is necessary to develop a new methodology for characterizing, rating and labeling shading fabric products, considering their relative performance in terms of energy use, daylight provision and visual comfort. (C) 2016 Elsevier B.V. All rights reserved.