The human A(3) adenosine receptor agonistic activity of 2-chloro-N(6)-substituted-4'-thioadenosine-5'-uronamides has been analyzed through Combinatorial Protocol in Multiple Linear Regression (CP-MLR) using 488 topological descriptors obtained from DRAGON software for the energy minimized 3D-structures of these molecules. Among the various descriptor classes considered in the study, the human A(3) adenosine receptor agonistic activity is correlated with simple topological descriptors (TOPO), Modified Burden eigenvalues (BCUT) and functional group (FUNC) classes of descriptors. The average valence connectivity index of order zero, X0Av, the sum of topological distances between O and Cl, T (O...Cl) from the TOPO class, the lowest eigenvalue n.2 of Burden matrix/weighted by atomic masses, BELm2, from the BCUT class and the number of secondary aliphatic amides, nCONHR, from FUNC class have contributed significantly in the development of a statistical sound models. The models developed and the participating descriptors suggest that the substituent groups at N(6)-position and/or 5'-uronamide of 2-chloro-N(6)-substituted-4'-thioadenosine-5'-uronamide derivatives hold scope for structural modification in the optimization of activity.