BACKGROUND\nVascular adhesion protein-1 (VAP-1) associates to subclinical atherosclerotic manifestations in young people, but its association to incident major adverse cardiovascular events (MACEs) and cardiovascular mortality in a general population is not known.\nMETHODS AND RESULTS\nWe used a newly developed ELISA to measure soluble VAP-1 (sVAP-1) levels in 2775 participants (mean age, 60 years) from a prospective cohort study (the FINRISK 2002). During a mean follow-up of 9 years, 265 participants underwent a MACE, and these participants had higher levels of sVAP-1 than those without MACE (868 ng/mL and 824 ng/mL, respectively, P<0.001). In multivariate-adjusted Cox proportional hazard model including traditional Framingham risk factors (age, sex, systolic blood pressure, cholesterol, high-density lipoprotein cholesterol, smoking, prevalent diabetes mellitus, and antihypertensive treatment), sVAP-1 independently predicted incident MACE (P=0.0046) and MACE mortality (P=0.026). The impact of sVAP-1 in predicting the 9-year absolute risk of MACE was analyzed using integrated discrimination improvement and net reclassification improvement with 10-fold cross-validation. Inclusion of sVAP-1 in the Framingham model improved integrated discrimination improvement (P=0.042), and the clinical net reclassification improvement by correctly reclassifying 9% (P=0.0019) of people in the intermediate risk (5%-20%) group.\nCONCLUSIONS\nsVAP-1 associated with increased risk of MACE and MACE mortality in people aged >50 years without prior MACE, and inclusion of sVAP-1 in the risk prediction model improved the clinical net reclassification improvement of incident MACE. Thus, sVAP-1 may be a potential new biomarker for cardiovascular diseases.