A numerical study including a validation process with experimental data was performed on the forced convection flows of nanofluids; the object of study was water containing TiO2 nanoparticles in smooth and micro-fin tubes at a constant wall temperature. Constant heat flux and temperature-dependent properties were used to determine the hydrodynamics and thermal behaviors of the nanofluid flow; a single-phase numerical model was used to solve two-dimensional equations by means of a CFD program for the water flow, contained in a smooth tube and in various micro-fin tubes having various helix angles (0 degrees, 18 degrees). An extensive literature review on the determination of the physical properties (k, mu, rho, Cp) of nanofluids is given in this paper. Multilayer Perceptron (MLP), one of theartificial neural network (ANN) methods, was used to determine the most agreeable physical propertiesof TiO2 nanofluids among correlations. The inputs ofthe ANN analyses were the correlations of physical properties, the average temperature and velocity of water in the test tubes, and the nanoparticle concentrations, while the outputs were shear stress, friction factor, heat flux, convective heat transfer coefficient, and pressure drop. After obtaining the best combination of physical properties of TiO2 nanofluids from the ANN analyses, the numerical model was validated by means of a CFD program, with the experimental smooth tube data as a case study; it was also validatedas a simulation studyfor several micro-fin tubes through a CFD program. This paper shows temperature, pressure, and velocity distributions in the investigated tubes; in addition, average and local experimental, theoretical, and numerical values in the smooth and micro-fin tubes are compared with oneanother in terms of friction factors, shear stresses, convective heat transfer coefficients, and pressure drops, according to various nanoparticle concentrations.