In this study, a method is used to develop a model for estimating luminance distributions of Thai's sky. From the sky characterization, five out of the fifteen CIE standard skies are selected for the distribution estimation. This selection represents the minimum sky types that best fit to the distribution records of Bangkok (Thailand) and cover three sky conditions of clear, partly cloudy, and overcast. The developed model uses the five standard skies each to estimate prevailing luminance distributions within a particular range of insolation condition defined by lower and upper limit values of Perez's clearness (e) and sky brightness (d). In this scheme, a stochastic search technique of evolutionary computation was applied to optimize limit values for an accurate sky classification. Equations of optimal limit values were also formulated as part of the method. From the comparative evaluation with other sky models (i.e. Perez's model, ASRC-CIE model, Harrison's model, CIE-Si model, and Igawa's all-sky model), the proposed model offers an alternative option for generating luminance distribution of Thai's sky. (C) 2016 Elsevier Ltd. All rights reserved.