An algorithm is presented for the generation of a reliable peptide component peak table from liquid chromatography-mass spectrometry (LC-MS) and subsequent quantitative analysis of stable isotope coded peptide samples. The method uses chemical noise filtering, charge state fitting, and deisotoping toward improved analysis of complex peptide samples. Overlapping peptide signals in mass spectra were deconvoluted by correlation with modeled peptide isotopic peak profiles. Isotopic peak profiles for peptides were generated in silico from a protein database producing reference model distributions. Doublets of heavy and light labeled peak clusters were identified and compared to provide differential quantification of pairs of stable isotope coded peptides. Algorithms were evaluated using peptides from digests of a single protein and a seven-protein mixture that had been differentially coded with stable isotope labeling agents and mixed in known ratios. The experimental results correlated well with known mixing ratios.