Abstract Alterations in mitochondrial dynamics, including their trafficking, can present early manifestation of neuronal degeneration. However, current methodologies used to study mitochondrial trafficking events rely on parameters that are mostly altered in later stages of neurodegeneration. Our objective was to establish a reliable computational methodology to detect early alterations in neuronal mitochondrial trafficking. We propose a novel quantitative analysis of mitochondria trajectories based on innovative movement descriptors, including straightness, efficiency, anisotropy, and kurtosis. Using biological data from differentiated SH-SY5Y cells treated with mitochondrial toxicants 6-hydroxydopamine and rotenone, we evaluated time and dose-dependent alterations in trajectory descriptors. Mitochondrial movement was analyzed by total internal reflection fluorescence microscopy followed by computer modelling to describe the process. The stacks of individual images were analyzed by an open source MATLAB algorithm ( www.github.com/kandelj/MitoSPT ) and to characterize mitochondria trajectories, we used the Python package trajpy ( https://github.com/ocbe-uio/trajpy/ ). Our results confirm that this computational approach is effective and accurate in order to study mitochondrial motility and trajectories in the context of healthy and diseased neurons in different stages.