A growing number of literatures appealed the necessity to develop methods of data processing for peptidome profiling and analysis. Although some methods had been established, many of them focused on the development and application of auto integration softwares. In this work, we paid attention to comparison of auto integration by software and manual integration for peptidomics data based on high performance liquid chromatography coupled with mass spectrometry (HPLC-MS). Two data processing procedures, auto integration by XCMS and manual integration, were applied in processing of peptidomics data based on HPLC-MS from cerebral infarction and breast cancer patients blood samples, respectively. And, it was found that almost all peaks contained in chromatograms could be picked out by XCMS, but the areas of these peaks were greatly different from those given by manual integration. Furthermore, t-test (2-tailed) results of the two data processing procedures were also different and different potential biomarkers were obtained. The results of this work will provide helpful reference for data processing of peptidomics research.