To evaluate the causal relationship between genetically determined telomere length (TL) and atherosclerosis (AS). We performed a 2-sample Mendelian randomization (MR) study to assess the potential causal relationship between TL and AS (coronary AS, cerebral AS, peripheral atherosclerosis (PAD), and AS, excluding cerebral, coronary, and PAD). The TL phenotype contained 472,174 participants, and the 4 subtypes of AS had 361,194, 218,792, 168,832, and 213,140 participants, all of European ancestries. The single nucleotide polymorphisms (SNPs) of TL strongly associated with the 4 atherosclerotic subtypes included in this study were 101, 92, 91, and 92, respectively. The odds ratios (ORs) and 95% confidence interval (CI) between TL and coronary AS calculated using inverse variance weighted (IVW) were 0.993 (0.988, 0.997), and the results were statistically significant (P < .05). The results between TL and cerebral AS, PAD, and AS (excluding cerebral, coronary, and PAD) were not statistically significant (P > .05). “Egger-intercept test” showed that there was no horizontal pleiotropy (P > .05); “leave-one-out analysis” sensitivity analysis showed that the results were stable and there were no instrumental variables with strong effects on the results; “MR- pleiotropy residual sum and outlier (PRESSO) test” showed 1 outlier for coronary AS and no outliers for the remaining subgroups. The results of the 2-sample MR analysis showed a causal association between TL and coronary AS but not with cerebral AS, PAD, and AS (excluding cerebral, coronary, and PAD). This may elucidate the observation that various vascular regions can be affected by AS but highlights the propensity of coronary arteries to be more susceptible to AS development.