Non-steroidal anti-inflammatory drug (NSAID) induced gastrointestinal toxicity has attracted greater attention over the years. The development of NSAIDs having safer therapeutic profile depends on the better understanding of their mechanisms, physicochemical and pharmacokinetic properties. The present investigation is aimed at in silico three dimensional quantitative structure-pharmacokinetic relationship (3D-QSPkR) assessment of a group of NSAIDs using self-organizing molecular field analysis (SOMFA) approach. Two different statistically validated models for dissociation constant (pK(a)) and volume of distribution (V(d)) were obtained from SOMFA studies of 22 clinically active NSAIDs having diversity in structure. 3D-QSPkR models delivered useful information about the contribution of shape and electrostatic potential on pK(a) and V(d). The study illustrated the significance of structural variables in molecular architecture of NSAIDs especially etodolac for further optimization of ADMET properties with improved therapeutic profile.