Objective: To develop standard experimental methods to minimise technical variance in matrix preparation for MALDI-TOF MS (matrix-assisted laser desorption/ionisation time-offlight) profiling. Methods: Matrix factors in saliva and serum samples of 20 healthy volunteers were examined, assuring their peptide components using seven different matrix type/preparation methods, HCCA(a-cyano-4-hydroxycinnamic acid)/SM(sample/matrix), SA(sinapinic acid)/DD(dried droplet), SA/SM, DHB(2.5-dihydroxyhenz-zoic acid)/DD, DHB/SM, DHAP(2.5-dihydroxyacetophenone)/ DD, DHAP/SM. Number of peaks, S/N(signal to noise) ratio and approximate range of target peaks were set as main selection criteria to find if these spell out any common regularity in results. Results: Different methods perform differently. DHB/DD performed worst in both samples, with no effective peak detected. For saliva sample, the S/N ratios of other six methods were lower. M/z range distributed differently. DHB/SM and DHAP/DD performed best in number of peaks, m/z distributing in 1000 to 2000 account for the vast majority. For serum sample, S/N ratios and m/z range distribution were different in different methods. S/N ratio of SA/DD and SA/SM were higher, number of peaks and m/z distribution were not irreplaceable. S/N ratios of the other four methods were lower. Conclusion: DHAP/DD and HCCA/SM performed best in number of peaks, m/z in 5000 - 7000 account for the vast majority in HCCA/SM and 1000 - 2000 in DHAP/DD. Further studies should focus on other characteristics of peptide components detected in different matrix methods to increase evidence when selecting matrix type/preparation methods. Palabras clave: MALDI-TOF MS, matrix preparation, proteomics, peptides, saliva