This study represents the performance of Fe2O3/water nanofluid (20 nm) in a double-pipe heat exchanger equipped with twisted-tape inserts for enhancing heat transfer. Fe2O3/water nanofluid is used because it has a higher thermal conductivity. Considered parameters include mass flow rate, twist ratio of tape, temperature and the volumetric portion of nanoparticles to water. The nanoparticles volume concentrations are 0.08% and 0.1% (v/v), and different twisted tapes with twist ratios of 2.5 ≤ y/w ≤ 5.2 are used. The Reynolds number variation is within the turbulent flow regime of 5000 < Re < 28,500. The obtained results show that adding nanoparticles as well as twisted-tape inserts increases heat transfer and Nusselt number. However, the effects of nanoparticles are more pronounced in high Reynolds number flows. Combining the positive effects of nanofluid and twisted tape, Nusselt number is significantly improved up to 103.45% in the test case. Moreover, there is no major change in friction factor. A multilayer perceptron (MLP) artificial neural network (ANN) with Levenberg–Marquardt (LM) learning algorithm and tangent sigmoid nonlinear transfer function is implemented for the aim of modeling the Nusselt number.