Stability of vehicle handling is of great importance for measuring the safety of vehicles. According to the current evaluation method, this crucial property is usually determined by technicians using existing impact factors and the evaluation is always generated by subjective judgments because there are no united criterion that can simplify the evaluation. However, this evaluation process is very hard to achieve because of the large scale of independent variables. Here, we aim at presenting a novel method based on artificial neural networks (ANNs) to aid the evaluation process for the tests of stability of vehicle handling. We set different impact factors of the tests as the independent variables, while the scores of the tests were set as the dependent variables. Using the existing data, we trained it using linear predictor, general regression neural network (GRNN) and multi-layer feedforward network (MLFN) during the machine learning process. Results show that ANN models can be used for aiding the subjective evaluation of stability in vehicle engineering. Our research can offer a novel insight for the vehicle evaluation in future studies.