Midregional Pro-enkephalin A (MR-PENK A) and N-terminal Protachykinin A (NT-PTA) have been associated with vascular dementia. However, the longitudinal relationship between these biomarkers and incident dementia has not been fully investigated. In the population-based Malmö Preventive Project, circulating levels of MR-PENK A and NT-PTA were determined in a random sample of 5,323 study participants (mean age: 69 ± 6 years) who were followed-up over a period of 4.6 ± 1.6 years. The study sample included 369 patients (7%) who were diagnosed in the same period with dementia. We analyzed relationship of MR-PENK A and NT-PTA with the risk of developing dementia by using multivariable-adjusted Cox regression models adjusted for traditional risk factors. Increased plasma levels of MR-PENK A were associated with higher risk of incident vascular dementia whereas no associations were found with all-cause or Alzheimer dementia. The risk of vascular dementia was mainly conferred by the highest quartile of MR-PENK as compared with lower quartiles. Elevated levels of NT-PTA yielded significant association with all-cause dementia or dementia subtypes. Elevated plasma concentration of MR-PENK A independently predicts vascular dementia in the general population. MR-PENK A may be used as an additional tool for identifying vascular subtype in ambiguous dementia cases.