Proteomic studies on mouse brain protein expression are still holding center stage as the generation of a reference database for the brain proteome, a need for designing expressional studies at the protein level. We therefore decided to extend the amount of identified brain proteins by the use of prefractionation. In order to reduce the complexity of mouse brain proteome we applied chromatographic prefractionations, ion-exchange and hydrophobic interaction chromatography, prior to 2-DE, followed by mass spectrometric identification (2-DE MALDI-MS). We analyzed about 17 000 protein spots in cytosolic fractions of mouse brain and identified about 10 000 spots. A total of 1841 proteins showing different pl or M-p representing probably post-translational modifications or splice variants, were products of 789 different genes. Numerous proteins were clearly identified as metabolic, antioxidant, cytoskeleton, signaling, transcription/translation, nucleic acid-binding, proteolysis-related proteins. We additionally provided evidence for the existence of hypothetical proteins predicted from nucleic acid sequences. Moreover, observed pls of proteins are listed thus enabling localization of proteins in a gel, information that cannot be obtained from theoretical pl's in databases. The results represent so far the largest database of mouse brain proteins and provide valuable information for the design of proteomic studies in the mouse.