BACKGROUND AND AIM: Insomnia has been implicated in gastrointestinal diseases (GIs), but the causal effect between insomnia and GIs and underlying mechanisms remain unknown. METHODS: By using the released summary-level data, we conducted a two-step Mendelian randomization (MR) analysis to examine the relationship between insomnia and four GIs and estimate the mediating role of candidate mediators. The first step was to investigate the causal association between insomnia and GIs using univariable MR analysis. The second step was to estimate the mediation proportion of selected mediators in these associations using multivariable MR analysis. Subsequently, results from different datasets were combined using the fixed-effect meta-analysis. RESULTS: Univariable MR analysis provided strong evidence for the causal effects of insomnia on four GIs after Bonferroni correction for multiple comparisons, including peptic ulcer disease (PUD) (odds ratio [OR] = 1.15, 95% interval confidence [CI] = 1.10-1.20, P = 1.83 × 10), gastroesophageal reflux (GORD) (OR = 1.19, 95% CI = 1.16-1.22, P = 5.95 × 10), irritable bowel syndrome (IBS) (OR = 1.18, 95% CI = 1.15-1.22, P = 8.69 × 10), and inflammatory bowel disease (IBD) (OR = 1.09, 95% CI = 1.03-1.05, P = 3.46 × 10). In the mediation analysis, body mass index (BMI) and waist-to-hip ratio (WHR) were selected as mediators in the association between insomnia and PUD (BMI: mediation proportion [95% CI]: 13.61% [7.64%-20.70%]; WHR: 8.74% [5.50%-12.44%]) and GORD (BMI: 11.82% [5.94%-18.74%]; WHR: 7.68% [4.73%-11.12%]). CONCLUSIONS: Our findings suggest that genetically instrumented insomnia has causal effects on PUD, GORD, IBS, and IBD, respectively. Adiposity traits partially mediated the associations between insomnia and GIs. Further clinical studies are warranted to evaluate the protective effect of insomnia treatment on GIs.