A series of sequential data processing steps were presented to support automatic protein identification by improved data quality of MSMS peptide fragment spectra. Proteins were digested according to Shevchenko et al. Peptide mixtures were separated by reversed phase HPCL at a flow rate of 200 nl/min. Results show that the quality of MSMS fragment spectra data improved by performing a sequence of data processing steps. Combination into one spectrum removed the bad quality spectra without a loss of information.