Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease of the central nervous system. It remains unclear whether pathological changes in ALS can lead to abnormalities in neural dynamics and how these abnormalities relate to key clinical characteristics of ALS. Nonlinear neural dynamics analyses of electroencephalography (EEG) sensorimotor channels were conducted using recurrence quantification analysis (RQA), permutation Lempel-Ziv complexity (PLZC), shannon entropy (ShannonE) and permutation entropy (PermEn). Whole-brain spatiotemporal topological dynamics were assessed using microstate analysis. The study shows that the nonlinear neural dynamics across all frequency bands in the sensorimotor channels of ALS patients are impaired (reduction in Shannon Entropy of Diagonal Line Length Distribution [ENTR]). Frequency-specific nonlinear neural dynamics indicate increased nonlinear neural dynamics in high-frequency bands (with increases in PLZC in beta2 (20-25 Hz)). Nonlinear neural dynamics in low-frequency bands decreases (with decreases in ShannonE in theta (4-7 Hz)) and is negatively correlated with disease duration. ENTR across all frequency bands and ShannonE in the theta band of the sensorimotor channels are potential protective factors for ALS. Furthermore, sensorimotor channel analysis shows a close relationship with whole-brain spatiotemporal topological neural dynamics. Duration A is positively correlated with RR, DET and ENTR, while Occurrence B is negatively correlated with it. The study demonstrates widespread abnormalities in the neural dynamics of ALS, which are closely related to clinical characteristics of ALS. There is also a close relationship between the neural dynamics of sensorimotor channels and whole-brain spatiotemporal topological dynamics in ALS patients.