Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder that is affected by both genetic and environmental factors. Nowadays, OMIC technologies, such as genomics and metabolomics, are providing a systematic readout of genetic structures and physiological states for understanding human diseases. However, the comprehensive analysis of cross-omics is often lacking. Here, we conducted a Mendelian randomization analysis to provide a comprehensive analysis of metabolomics and genomics to estimate the causal relationships between non-targeted human serum metabolites and the development of ALS. Using genetic variants as predictors, our study detected 18 metabolites that might have causal effects on the development of ALS, including a group of gamma-glutamyl amino acids. Our findings suggested that glutathione metabolism dysfunction might be involved in the pathogenesis of ALS. Furthermore, our study provides a novel method to understand the pathogenesis of human diseases and develop therapeutic strategies for diseases by combining metabolomics with genomics.