Because nearly all existing glare equations depend on source size, relative position, and luminance, predicting glare in daylighting software generally requires the pixel-processing of a rendered image meant to reproduce a human’s view. When considering multiple positions, views, and times of day and year, making an annual assessment of glare becomes prohibitively time consuming using traditional methods. Using the recently developed metric Daylight Glare Probability as a reference, this paper builds upon an existing approximation method for DGP based on vertical illuminance. The existing method performs well for glare situations based on high vertical illuminance but is less accurate for luminance contrast-based glare. The approach that is presented here suggests a way to overcome these weaknesses by using geometric information and illuminance data generated from the computer model.