Over-reporting of unreliable protein identifications has reduced the accuracy and reproducibility of MS/MS-based proteomic studies. In this work, we demonstrate the analysis workflow used by a bioinformatic tool called Scaffold, which attempts to increase the confidence in protein identification reports through the use of several statistical methods. In addition, this work describes an advanced protein grouping method used by Scaffold to further reduce falsely reported protein identifications, particularly when using large or otherwise sequence redundant protein databases.