BACKGROUND The physiological role of lipoprotein(a) is unclear; however, lipoprotein(a) may play a role in hemostasis and wound healing. We tested the hypothesis that high lipoprotein(a) concentrations are associated with low risk of major bleeding in the brain and airways both observationally and causally (from human genetics). METHODS We examined 109169 individuals from the Copenhagen City Heart Study and the Copenhagen General Population study, 2 similar prospective studies conducted in the Danish general population. Individuals had information on plasma lipoprotein(a) concentrations (n = 59980), LPA kringle-IV type 2 (KIV-2) number of repeats (n = 98965), and/or LPA single-nucleotide polymorphism rs10455872 associated with high lipoprotein(a) concentrations (n = 109 169), and information on hospital contacts or death due to major bleeding in brain and airways from registers. RESULTS Using extreme phenotypes or genotypes, the multifactorially adjusted hazard ratio for major bleeding in the brain and airways was 0.84 (95%CI: 0.71-0.99) for lipoprotein(a), >800 mg/L vs <110 mg/L; 0.83 (0.73-0.96) for KIV-2, <24 vs >35 number of repeats; and 0.89 (0.81-0.97) for rs10455872 carriers (heterozygotes + homozygotes) vs noncarriers. The corresponding hazard ratios were 0.89 (0.82-0.98) for heterozygotes and 0.59 (0.36-0.98) for homozygotes separately vs rs10455872 noncarriers. Also, for a 1 standard deviation higher lipoprotein(a) (= 310 mg/L), the hazard ratio for major bleeding in the brain and airways was 0.95 (95%CI: 0.91-1.00) observationally, 0.89 (0.80-0.98) causally based on LPA KIV-2 number of repeats, and 0.94 (0.87-1.02) causally based on LPA rs10455872. CONCLUSIONS High lipoprotein(a) concentrations were associated with lower risk of major bleeding in the brain and airways observationally and causally. This indicates that lipoprotein(a) may play a role in hemostasis and wound healing.