Background Cheese intake has been observed to be associated with obesity, but whether this relationship was causal was still unclear. To evaluate whether cheese intake have an causality effect on obesity and related traits. Methods We performed a two-sample Mendelian randomization (MR) analysis. Summary statistics from published Genome-wide association study (GWAS) associated with obesity and related traits including body mass index (BMI), waist circumference (WC), hip circumference (HC), waist-to-hip ratio (WHR) were obtained from FinnGen Biobank, IEU OpenGWAS database and UK Biobank. The cheese intake summary dataset was obtained from a UK Biobank database including 451,486 participants. Effect estimates were evaluated using inverse-variance weighted (IVW) method. Other MR analyses, including MR Egger (bootstrap), weighted median and weighted mode methods were performed in sensitivity analyses. Results In the MR analysis, genetically predicted cheese intake were associated with a 60% decreased risk of obesity (95% CI: 0.25, 0.63, P = 9.76E-5), 72% of obesity class 1 (95% CI: 0.16, 0.51, P = 1.75E-5), 74% of obesity class 2 (95% CI: 0.13, 0.55, P = 4.12E-4), 79% of obesity class 3 (95% CI: 0.07, 0.64, P = 6.25E-3) and 52% of overweight (95% CI: 0.33, 0.71, P = 2.39E-4) per 1-standard deviation (SD) increment of cheese intake by IVW. The results also provided strong evidence that cheese intake was causally and negatively associated with BMI (β=-0.60, 95% CI: -0.90, -0.29; P = 0.000), WC (β=-0.28, 95% CI: -0.40, -0.17; P = 2.13E-6), HC (β=-0.27, 95% CI: -0.41, -0.13; P = 2.13E-4) and WHR (β=-0.39, 95% CI: -0.50, -0.27; P = 5.92E-11). Results were robust using other MR methods in sensitivity analyses. Conclusions This two-sample MR analysis found that genetically predicted cheese intake had causal effect on an decreased risk of obesity and related traits. The improvement of dietary behavior may be beneficial to prevent obesity.