Basal cell-specific, anti-high molecular weight cytokeratin (HMCK) antibody is often used to confirm the diagnosis of prostatic adenocarcinoma, particularly if limited amounts of tissue are available. HMCK is formalin sensitive and requires pretreatment by enzymes or heat if formalin-based fixatives are used. To date, the effect of prolonged formalin fixation on HMCK immunoreactivity has not been systematically studied; this is critical, because the diagnosis of malignancy is based on a negative immunoreaction. In this study, 5 tissue blocks obtained from each of 10 radical prostatectomy specimens were fixed in formalin from 6 hours to 1 month. HMCK immunostaining was performed with monoclonal antibody clone 34betaE12 after pretreatment of the sections by either enzymatic predigestion with pepsin, heat-induced epitope retrieval (HIER) with a microwave, or HIER with a hot plate. For scoring, the staining intensity at 6 hours of formalin fixation was considered as the baseline for that particular antigen retrieval technique. After pepsin predigestion or microwaving, there was progressive loss of HMCK immunoreactivity from 1 week or longer of formalin fixation. HIER with a hot plate yielded consistent results with no decrease in HMCK immunoreactivity with as long as 1 month of formalin fixation. The staining intensity was consistently stronger at all periods of formalin fixation when the hot plate method was used, compared with pepsin predigestion or microwaving. Generally weak HMCK positivity was observed in rare neoplastic cells of 3 of 10 specimens after hot plate HIER but not with pepsin predigestion or microwave antigen retrieval. This sporadic immunostaining of malignant cells was quantitatively and qualitatively distinct from the pattern seen in benign epithelium. We conclude that formalin fixation affects HMCK immunoreactivity over time and might impact its diagnostic usefulness. Efficacies of different antigen unmasking/epitope retrieval techniques vary and must be standardized for individual laboratories.