A key issue in proteomics is to quantify changes in protein levels in complex biological samples under different conditions. Traditional two-dimensional gel (2-DE) electrophoresis-based proteomic approaches are tedious and suffer from several limitations, including difficulties in detecting low abundant and insoluble proteins. Isotope-coded affinity tagging (ICAT), one of the most employed chemical isotope labeling methods, can address many of the shortcomings of 2-DE. ICAT relies on the sensitivity of mass spectrometry (MS) to quantify relative protein abundance in a mixture of two differentially labeled protein samples. We describe here a detailed protocol for ICAT-based quantification of proteins in two or more biological samples, including sample preparation, ICAT labeling, fractionation and purification, and analysis by MS. For the MS analysis, we describe a "targeted" approach, which includes quantification of the samples using MS followed by selective identification of only the differentially expressed ICAT pairs using tandem MS (MS/MS). This approach gives more biologically relevant information than a data-dependent MS/MS analysis. We also describe the steps in data analysis, statistical analysis, and protein database searching.