Previous observational studies revealed a potential but partially controversial relation between lipid metabolism and risk of amyotrophic lateral sclerosis (ALS), potentially prone to bias. Therefore, we aimed to study whether lipid metabolism involves genetically determined risk factors for ALS through Mendelian randomization (MR) analysis. Using GWASs summary-level data for total cholesterol (TC) (n=188,578), high-density lipoprotein cholesterol (HDL-C) (n=403,943), low-density lipoprotein cholesterol (LDL-C) (n=440,546), apolipoprotein A1(ApoA1) (n=391,193), apolipoprotein B (ApoB) (n=439,214), and ALS (12,577 cases and 23,475 controls), we implemented a bidirectional MR study to evaluate a genetic relation between lipids and ALS risk. We performed a mediation analysis to assess whether LDL-C is a potential mediator on the pathway from traits of LDL-C-related polyunsaturated fatty acids (PUFAs) to ALS risk. We identified genetically predicted increased lipid levels to be associated with the risk of ALS, whereby elevated LDL-C had the most potent effect (OR, 1.028; 95% CI, 1.008-1.049, p=0.006). The effect of increased levels of apolipoproteins on ALS was similar to their corresponding lipoproteins. ALS did not cause any changes in lipid levels. We found no relation between LDL-C-modifying lifestyles and ALS. The mediation analysis revealed that LDL-C could act as an active mediator for linoleic acid, with the mediation effect estimated to be 0.009. We provided high-level genetic evidence verifying the positive link between preclinically elevated lipid and ALS risk that had been described in previous genetic and observational studies. We also demonstrated the mediating role of LDL-C in the pathway from PUFAs to ALS.