Purpose Low back pain (LBP) is a common health problem in the global population. This study aims to assess whether smoking initiation, alcohol consumption, and coffee consumption are causally with an increased risk of LBP. Methods A two-sample Mendelian Randomization (MR) study was designed, based on summary-level data from the largest published genome-wide association studies. Single nucleotide polymorphisms with genome-wide significance level (P < 5.0 x 10(-8)) were selected as instrumental variables for each exposure. Standard inverse-variance weighted (IVW) method was used as the primary statistical method. The weighted median, MR-Egger regression, and MR-PRESSO methods, which relax some IV assumptions, were used for sensitivity analysis. Results Genetically predicted smoking initiation was causally associated with higher odds of LBP. The pooled OR of LBP using IVW method was 1.36 (95%CI 1.22 1.52; P = 6.0 x 10(-8)) for one SD increase in the prevalence of smoking initiation, which was supported by the weighted median method (OR: 1.41, 95%CI 1.22, 1.64; P = 5.7 x 10(-6)). Sensitivity analysis confirmed the robustness of pooled OR of LBP. There was no evidence to suggest a causal effect of alcohol and coffee consumption on LBP. The pooled ORs of LBP were 1.36 (95%CI 0.94, 1.97; P = 0.10) for alcohol consumption and 1.00 (95%CI 0.99, 1.00; P = 0.17) for coffee consumption, respectively. Conclusion Smoking is casually associated with an increased risk of LBP. Smoking control should be recommended in LBP patients to avoid worsening the disease. The safety of LBP with moderate alcohol and coffee consumption merits more study.