Background: Coumarins are secondary metabolites from the phenylpropanoid-type biosynthesis in higher plants. A plethora of potential phytopharmacological activities have been described for derivatives of the coumarin scaffold: hepatoprotective, antineoplastic, antimicrobial, antituberculosis, antiviral, anti-inflammatory anticoagulant, or antithrombotic effects. Objective: A computer-based quantitative structure – activity relationships (QSAR) study for a series of 4?chloro- 3-formylcoumarins was carried out. Methods: To this end we generated the 3D models of 17 published coumarin structures, calculated their physicochemical properties (descriptors) to correlate them to their experimentally known biological activities measured as inhibition concentrations to block the target enzyme activity. Our proposed approach used free molecular modeling software and applies our scripts written in the programming language R. Results: The final multiple regression models achieved satisfactory results with a small number of descriptors– all of which were statistically significant and meaningful in the field of pharmacodynamics to develop new 3-formylcoumarins with enhanced activities targeting the human thymidine phosphorylase enzyme. Conclusion: On theoretical grounds, our in silico research contributes in a crucial step in the field of complementary phyto-medicine. This step is located between in vivo pharmacological observations of plant extracts on ethnopharmacological, preclinical or controlled clinical levels and the need to identify – at an atomic scale – all those plant ingredients responsible for the biological actions under scrutiny. Our simulations shed light on the modification of phyto-medicine’s physicochemical properties to enhance the interaction with their biomolecular target in the patient’s body.