Preanalytical sample handling and storage procedures play an extremely important role in reliably measuring neurochemical dementia diagnostics (NDD) biomarkers: Aβ(1-40), Aβ(1-42), Tau, and pTau181. To test different handling and storage conditions, the following protocols were applied: (a) storage at room temperature for one week, (b) deep-freezing and thawing up to three cycles, (c) deep-freezing, thawing and keeping under +4°C for two days before the analysis, and (d) long-term stability of a deeply frozen sample. Between the first and the seventh day of the storage at room temperature, the percentage of the concentrations (compared to the starting concentrations) fluctuated: 104.3-105.3, 97.6-93.2, 100.6-96.8, and 97.9-90.2 for Aβ(1-40), Aβ(1-42), Tau, and pTau181, respectively. Re-freezing cycles resulted in the percentage fluctuations of the concentrations: 101.1-105.5, 95.4-99.7, 98.3-100.0, and 100.5-101.4 for Aβ(1-40), Aβ(1-42), Tau, and pTau181, respectively. Keeping previously frozen/thawed samples under +4°C for two days resulted in the percentage differences of the concentrations: +15.9, +2.2, -1.1, and -0.1 for Aβ(1-40), Aβ(1-42), Tau, and pTau181, respectively. During long-term stability, the coefficients of linear correlation (R(2)) were: Aβ(1-40), 0.007; Aβ(1-42), 0.02; Tau, 0.011; and pTau181, 0.02, and the corresponding inter-assay coefficients of variation: 13.9%, 13.9%, 11.0%, and 10.7% for Aβ(1-40), Aβ(1-42), Tau, and pTau181, respectively. We conclude that the NDD biomarkers are relatively stable when the cerebrospinal fluid sample is kept at room temperature for about four days; one or two thawing/refreezing cycles do not profoundly affect the biomarkers concentrations, however three cycles result in increased unsystematic variation. The four biomarkers seem to be stable in a sample stored deeply frozen for more than two years.