Alzheimer's disease (AD) is the most frequent dementia characterized by formation of plaques and neurofibrilary tangles. Vascular dementia (VAD) is the second most frequent type of dementia, which is caused by ischemic, hypoperfusive or hemorrhagic brain lesions. The aims of this study are: assessment of potential serum biomarkers neurosin (human kallikrein 6, KLK6), clusterin (CLU), adiponectin (ADPN) and inflammatory marker interleukin – 6 (IL-6) in differential diagnostics of AD and VAD; assessment of potential of KLK6, CLU, ADPN and IL-6 to separate age-matched cognitively healthy individuals from those who are demented and to separate those individuals with symptoms of mild cognitive impairment (MCI) from those with overt AD or VAD; assessment of correlation of KLK6, CLU, ADPN and IL-6 with results of parameters regularly used for determination of cognitive deficit. 70 patients with diagnosis of AD and 67 patients with VAD were included in study in consecutive order during their routine neurological follow-up in University Department of Neurology in Medical School University Hospital Sestre milosrdnice, Zagreb. Control group of age-matched individuals consisted of cognitively healthy individuals (N = 50) and those with mild cognitive impairment (MCI) (N = 48). Neuropsychological tests Mini Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) were done as part of routine neurological examination. Neuroimaging techniques multi slice computed tomography (MSCT), Colour Doppler Flow Imaging (CDFI) and transcranial Doppler sonography were used. Routine biochemistry tests were determined on automated biochemistry analyzer. Concentrations of potential biomarkers were measured using ELISA method for KLK6 and CLU, and concentration of ADPN was measured using immunoturbidimetry on automated biochemistry analyzer. Concentration of IL-6 was determined using immunochemistry analyzer. Depending on data distribution, statistic tests Kruskal Wallis and ANOVA were used for difference testing. For correlation analysis Spearman and Pearson tests were performed. Statistical analysis was done with MedCalc program. Difference between tested groups was found for those tests which are routinely used for neurological follow-up of demented patients. Concentrations of potential biomarkers KLK6, CLU and ADPN did not differ between tested participant groups (P = 0,137, P = 0,178 and P = 0,268). Concentrations of IL-6 were significantly different for tested groups (P = 0,014), with highest median of concentration in VAD group (4,1 pg/mL) and lowest median of concentration in MCI group (2,3 pg/mL). Concentrations of investigated biomarkers KLK6, CLU and ADPN did not differ significantly between AD and VAD patient group. Also, capability of tested biomarkers to differentiate age-matched cognitively healthy participants from patients with diagnosis of dementia was not sufficient, as well as their discriminating potential for age-matched participants with MCI compared to AD and VAD patient groups. Measurement of concentration of inflammatory marker IL-6 proved to be useful in discriminating between tested groups. Concentrations of potential biomarkers did not correlate with results of standard indicators of cognitive deficiency.