A novel strategy, termed protein array pixelation, is described for comprehensive profiling of human plasma and serum proteomes. This strategy consists of three sequential high-resolution protein prefractionation methods (major protein depletion, solution isoelectrofocusing, and 1-DE) followed by nanocapillary RP tryptic peptide separation prior to MS/MS analysis. The analysis generates a 2-D protein array where each pixel in the array contains a group of proteins with known pI and molecular weight range. Analysis of the HUPO samples using this strategy resulted in 575 and 2890 protein identifications from plasma and serum, respectively, based on HUPO-approved criteria for high-confidence protein assignments. Most importantly, a substantial number of low-abundance proteins (low ng/mL - pg/mL range) were identified. Although larger volumes were used in initial prefractionation steps, the protein identifications were derived from fractions equivalent to approximately 0.6 microL (45 microg) of plasma and 2.4 microL (204 microg) of serum. The time required for analyzing the entire protein array for each sample is comparable to some published shotgun analyses of plasma and serum proteomes. Therefore, protein array pixelation is a highly sensitive method capable of detecting proteins differing in abundance by up to nine orders of magnitude. With further refinement, this method has the potential for even higher capacity and higher throughput.