In this paper, a neutral-path departure algorithm is proposed to define safe handling threshold conditions and dangerous steering situation for powered two-wheeled vehicles. Based on this study, a Self Steering Gradient for motorcycles is proposed as a risk function for neutral-path departure detection. Furthermore, the motorcycle overturning or under-steering are analyzed based on the handling index. This index depends on the intrinsic motorcycle parameters, as well as, the state outputs. The proposed neutral-path departure algorithm aims to assess the risk when the motorcycle begins to drift out of the neutral path. Finally, the effectiveness of the detection scheme is tested using a high-fidelity software BikeSim©.