A correction to the reported liquid–liquid equilibrium parameters using the non-random two liquid (NRTL) thermodynamic model for the ethanol/toluene/n-decane system at 298 K is reported. The parameters were calculated by minimising the residual between the calculated coexistence and the experimental compositions. However, to obtain a physically plausible coexistence curve, a parameter in the NRTL model had to be fixed. This highlights the importance of assessing the entire coexistence curve, as opposed to only comparing the calculated compositions to the experimental data points. This is because undertaking the regression for all the available parameters will result in a smaller residual and better fit of the calculated points to the experimental points. This leads to an apparent improved fit but the complete coexistence curve will show that the parameters are not physically plausible.