Abstract This paper presents a study on the effect of mass fraction of nanoparticles, cooling rates and prediction of latent heat of fusion for Barium Chloride Dehydrate solutions (BaCl2·2H2O) by separately adding a mass fraction of 0.2 wt%–1 wt% magnesium oxide (MgO) and multi-walled carbon nanotubes (MWCNTs) and various cooling rates applied. The data was then compared with existing prediction model and a new correlation developed. The results show that the latent heat of fusion reduced by 7% and 5.2% for MWCNT and MgO nanofluids respectively at a mass fraction of 1 wt% and at a cooling rate of 5 °C/min. Mass loss equation maximum deviation was 5.55% and 4.16% for MWCNT and MgO nanofluid respectively at a mass fraction of 1 wt% and at a cooling rate of 5 °C/min. The new correlation maximum absolute deviation was 1.2% at a mass fraction of 0.4 wt% and at a cooling rate of 5 °C/min for MWCNT nanofluid while for MgO nanofluid, the deviation was 1.7% at a mass fraction of 1 wt% and at a cooling rate of 10 °C/min confirming the accuracy of the new correlation and therefore can be applied to predict the latent heat of fusion of any nanofluid.