The advent of proteomics techniques has been enthusiastically accepted in most areas of biology and medicine. In neuroscience, a host of applications was proposed ranging from neurotoxicology, neurometabolism, determination of the proteome of the individual brain areas in health and disease, to name a few. Only recently, the limitations of the method have been shown, hampering the rapid spreading of the technology, which in principle consists of two-dimensional gel electrophoresis with in-gel protein digestion of protein spots and identification by mass-spectrometrical approaches or microsequencing. The identification, including quantification using specific software, of brain protein classes, like enzymes, cytoskeleton proteins, heat shock proteins/chaperones, proteins of the transcription and translation machinery, synaptosomal proteins, antioxidant proteins, is a clear domain of proteomics. Furthermore, the concomitant detection of several hundred proteins on a gel allows the demonstration of an expressional pattern, rather generated by a reliable, protein-chemical method than by immunoreactivity, proposed by protein-arrays. An additional advantage is that hitherto unknown proteins, so far only proposed from their nucleic acid structure, designated as hypothetical proteins, can be identified as brain proteins. As to shortcomings and disadvantages of the method we would point to the major problem, the failure to separate hydrophobic proteins. There is so far no way to analyse the vast majority of these proteins in gels. Several other analytical problems need to be overcome, but once the latter problem can be solved, there is nothing to stop the method for a large scale analysis of membrane proteins in neuroscience.