Modeling of natural convection heat transfer in an inclined C-shape cavity is studied in this paper. The enclosure is filled with H2O-Fe3O4 nanofluid under the effect of magnetic field. The operating range of parameters used in this study were Hartmann number (Ha) from 0 to 80, Rayleigh number (R-a) from 1E2 to 1E6, nanoparticles volume fraction (phi) from 0 to 0.1, inclination angle (alpha) from 0 to 90 deg, and aspect ratio (AR) from 0.2 to 0.8. The employed model is solved using CFD tools based on the finite element method. The comparison with reference experimental data indicated the accuracy and generalization capability of the model. In addition, a novel correlation and an artificial neural network (ANN) model were productively developed for predicting Nu number as a function of aforementioned independent variables. The influence of the model parameters on the Nu number is precisely presented and discussed. It is shown that Ra number and aspect ratio have more impact on Nu than the other variables. (C) 2017 The Society of Powder Technology Japan. Published by Elsevier B.V. and The Society of Powder Technology Japan. All rights reserved.