Potent and selective ligands with a nucleoside skeleton are generally thought as agonists of the human A(3) adenosine receptor (AR), however, some of them can also act as full antagonists. This work reports a Quantitative Structure-Activity Relationship (QSAR) study for predicting the binding affinity of such type of compounds towards the A(3) AR. Several different theoretical molecular descriptors, calculated only on the basis of knowledge of the molecular structure and an efficient variable selection procedure, such as forward stepwise regression, led to models with satisfactory accuracy and predictive ability. But the best-final QSAR model is based on the Molecule Representation of Structures based on Electron diffraction (3D-MoRSE) descriptors capturing a reasonable interpretation. This QSAR model is able to explain more than 85% of the variance in the experimental affinity and manifests good predictive ability as indicated by the higher Q(2)s of cross- and external-validations. The model obtained in this study may provide guidance for future design of new potent and selective human A(3) AR full antagonists with a nucleoside skeleton.