As the decision for immunocytochemistry is usually made on the basis of findings in Papanicolaou-stained smears and uncovering of the smears takes time, the immunocytochemical results are often reported with some delay. But they are of clinical interest only if reported within a short time. Therefore, immunocytochemistry on cytologic preparations must be carefully organized. The decision for immunocytochemistry must be made before the mounting medium has completely hardened to keep the time of uncovering short. The method of immunocytochemistry should fulfill the following prerequisites: 1. Cell sampling and fixation should be easy to handle for the clinician who sends the specimen to the laboratory. 2. Unspecific background staining, especially in cytologic preparations rich in blood and protein, should not occur. 3. The immunostaining method should be applicable to all kinds of cytologic material, fixed and stained smears included. 4. The nuclear structure of tumor cells should not be destroyed by the immunocytochemical procedure so that tumor cells after incubation are clearly distinguishable from normal cells showing a similar reaction as the tumor cells. There has hitherto been no such all-round method fulfilling all these prerequisites since the properties of the antigenic epitopes of the cells and of the antibodies recognizing them are too heterogeneous. Therefore several methods have to be considered and a variety of technical aspects such as fixation, storage of cytologic material, properties of tinctorial stains, of antibodies and of the antigenic epitopes must be studied to find out the two or three standard methods which meet the requirements in most cases. We recommend the ABC method for Papanicolaou-stained smears and the APAAP method for demonstration of lymphocyte markers. The indication of immunocytochemistry in diagnostic cytology is restricted by the limited number of specimens. Therefore, the following rules have to be observed: 1. The conventional light-microscopic examination must have priority over the immunocytochemical examination. 2. The cytologic specimens assigned for immunocytochemical examination must have been adequately fixed and stored. 3. As the number of smears is limited, the immunocytochemical examinations must be carefully planned and restricted to the absolutely necessary incubations. If possible, an informative smear has to be spared for documentation and future training of cytologists and cytotechnicians. 4. Immunocytochemical examinations in cytology are only justified if the diagnostic problem can be clearly defined. 5. The panel of antibodies should be selected carefully so that the results may give an answer to alternative questions. At least two antibodies should be applied.(ABSTRACT TRUNCATED AT 400 WORDS)