The significance of proteomic research is coupled with the recent exponential growth of these investigations. Currently, the most popular techniques used for these studies include the coupling of 1- and 2-dimensional electrophoresis with mass spectrometric analysis of the extracted and digested proteins. However, detection limits of gel staining methods have led to a search for complimentary techniques that afford the detection of lower concentrations of biologically relevant proteins. In the present studies, we have evaluated the applicability of on-line capillary electrophoresis - mass spectrometry (CE-MS) for this application. Specifically, we used membrane preconcentration-CE-MS (mPC-CE-MS) to analyze 13 samples of human aqueous humor (AH) from patients with various ocular pathologies (cataract, cataract plus glaucoma, and cataract plus pseudoexfoliation syndrome). This approach enabled rapid analysis of a relatively large volume (1 microL of each specimen, and a protein map for each was created. Measured average molecular weights (Mr) were used to tentatively identify proteins after search of the SWISS-PROT database using TagIdent from ExPaSy. Among those proteins tentatively identified are beta-2 microglobulin (Mr 11731.2), apolipoprotein A1 (Mr 28078.6) and serum albumin (Mr 66400). Proteins with Mr of 4349 (unidentified), 11731.2 (beta-2 microglobulin), 13400-14100 (immunoglobulin fragments), 28078.2 (apolipoprotein A1) and approximately 68000 (serum albumin) were observed in the majority of specimens. Generally no significant differences were noted in the protein composition of aqueous humor samples from different pathologies. However, the absence of an Mr 13345 protein and its oxidized form (Mr 13361) in samples from patients with pseudoexfoliation syndrome was noted. Occasionally the alpha-and beta-chains of hemoglobin, a contaminant in aqueous humor introduced during sampling, were also detected. We conclude from these studies that mPC-CE-MS is an attractive complimentary technique for proteome research, as this approach enables direct mapping and characterization of low concentrations of proteins that are present in complex physiologically derived fluids.