Antibodies have been in widespread use for more than three decades as invaluable tools for the specific detection of proteins or other molecules in biological samples. In spite of such a long experience, the field of immunocytochemistry is still troubled by spurious results due to insufficient specificity of antibodies or procedures used. The importance of keeping a high standard is increasing because massive sequencing of entire genomes leads to the identification of numerous new proteins. All the identified proteins and their variants will have to be localized precisely and quantitatively at high resolution throughout the development of all species. Consequently, antibody generation and immunocytochemical investigations will be done on a large scale. It will be economically important to secure an optimal balance between the risk of publishing erroneous data (which are expensive to correct) and the costs of specificity testing. Because proofs of specificity are never absolute, but rather represent failures to detect crossreactivity, there is no limit to the number of control experiments that can be performed. The aims of the present paper are to increase the awareness of the difficulties in proving the specificity of immunocytochemical labeling and to initiate a discussion on optimized standards. The main points are: (1) antibodies should be described properly, (2) the labeling obtained with an antibody to a single epitope needs additional verification and (3) the investigators should be required to outline in detail how they arrive at the conclusion that the immunocytochemical labeling is specific.