With a single-phase approach, a 3D mathematical model was developed through Computational Fluid Dynamics (CFD) techniques, coupling the momentum and heat transfer balances for the study of the thermal behaviour of nanofluids. The local heat transfer coefficient and thermal boundary layer thickness of CuO/water, Fe2O3/water and Al2O3/water nanofluids, have been predicted and compared with those experimentally determined at three volume concentration of nanoparticles (Ø=0.01%, 0.03% and 0.05%), at T = 25 °C and T = 55 °C for laminar and turbulent flow conditions, using a newly developed sophisticated noninvasive heat transfer coefficient probe that is flush mounted on the inner wall of the test section. Such conditions have been used for the mathematical model, considering the effects of the nanoparticle materials and volume concentrations through effective thermophysical properties. The predicted results from the mathematical model show a good agreement with the trend and the experimental observations. The enhancement of the heat transfer coefficient and reduction of the thermal boundary layer when increasing the volume concentration of the nanofluids and when increasing the flow rates have been properly predicted by the mathematical model results, showing average absolute relative errors between 1.7% and 8.4%. The model exhibits an enhancement in the agreement between the experimental measurements and the predicted results when comparing with other models found in literature.