Human sweet taste receptor structure is predicted and was further evaluate using experimental result. This would enable receptor-based docking and pharmacophore studies since no structure, experimental or predicted model, for complete subunits of human sweet taste receptor (hSTR) is available till date. Different homology modelling as well as threading-based tools were employed for structure prediction of individual domains (amino terminal domain (ATD), cysteine rich domain (CRD) and transmembrane domain (TMD)) and complete subunit structure. Finally, complete subunits structure was built from combinations of individual domain models. These predicted models were validated and further evaluated using docking and interaction analysis of the experimentally studied sweet molecules. hSTR Modelling through threading based tools was of poor quality with the exception of ITASSER software that predicted models with greater than 90% residues in energetically favourable environment. Among homology based software, CPH Model, SWISS Model and Prime predicted hSTR model of acceptable quality with more than 95% residues in energetically favourable environment. This model can be used for receptor-based pharmacophore modelling or searching newer sweet molecules.