Biodiesel has proven to be a feasible source of renewable energy, but the presence of bisallylic carbons within its structure renders it susceptible to oxidation. The employment of additives to mitigate the autooxidation of fuel is a solution, but modelling of the relationship between additive type and concentration with oxidation stability and related properties has yet to be developed for palm biodiesel. This paper aims to discuss the use of artificial neural network (ANN) to develop models for the said relationship to reduce experimental costs and time. Using MATLAB, ANN models for oxidation stability (Rancimat and PetroOXY), acid value and kinematic viscosity were developed. The mean squared error (MSE) and correlation (R2) values were used to validate the accuracy of the ANN models. The results showed that the Rancimat and PetroOXY models were the most accurate, while acid value and kinematic viscosity recorded lower accuracy possibly due to insignificant trends observed in the data set. It was found that 7–8 neurons in hidden layer resulted in the best predictive performance.