Immunocytochemical studies on fine needle aspiration (FNA) material often yield inconsistent results. We undertook this study to determine the influence of variation in specimen preservation and processing on results of immunocytochemical staining. Multiple cytospin slides were made from material obtained by FNA of 15 surgical specimens of carcinomas. Variables studied included transport medium/initial preservative (50% ethanol versus CytoLyt), time (CytoLyt or PreservCyt for up to 5 days), slide fixative (acetone, 95% ethanol, or alcohol: formalin 50: 50 vol: vol), and time in fixative (1, 5, 10, or 20 minutes). Slides were stained with a cocktail of antibodies recognizing cytokeratin. Formalin-fixed tissue sections of corresponding surgical specimens served as positive controls; cytospin slides stained with mouse myeloma protein served as negative controls. The stained slides were evaluated using a semi quantitative system for intensity of staining (i.e., 1, 2, or 3, in which 3 equals the same intensity as the tissue sections), background staining (clean versus dirty/obscuring), and cellular preservation (cell morphology: 1, poor; 2, fair; 3, good). CytoLyt and 50% ethanol are comparable in all three aspects (staining intensity, background, and cellular preservation) evaluated. Fixation with formol alcohol gave suboptimal immunocytochemical staining, whereas acetone and 95% ethanol are comparable in this regard. Five minutes appears to be the optimal duration for cellular preservation on slides fixed in acetone or 95% ethanol. Other than these findings, none of the procedural variations made a detectable difference in the results. Antigenicity was well preserved in cells maintained in CytoLyt or PreservCyt for up to 5 days.