Previously we observed that hyperplastic epidermal keratinocytes characteristic of psoriasis had abundant amounts of the cell survival protein Bcl-xL; however, whether this overexpression correlated with enhanced survival was unclear because the majority of epidermal cells possess nuclei that are positively labeled by an assay typically regarded as indicative of cells undergoing apoptosis (terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end-labeling (TUNEL) staining). To clarify this apparent discrepancy, we explored the propensity of keratinocytes derived from psoriatic plaques to undergo apoptosis and also determined the reliability of TUNEL staining as an indicator of apoptosis in keratinocytes in vitro and in vivo. First, a keratinocyte cell line, HaCat, was examined before and after being suspended in semisolid medium (methylcellulose) using flow cytometry to detect TUNEL-positive cells, and the percentage of positive cells was correlated to the presence or absence of double-stranded DNA fragmentation using pulsed field gel electrophoresis. After 18 hours in methylcellulose suspension, apoptosis was detected in HaCat cells when at least 5% of the cell population was undergoing programmed cell death. Second, we examined 23 clinical specimens of skin (13 from psoriatic patients and 10 from healthy control subjects) and observed that no double-stranded DNA fragmentation was present in any of the freshly isolated keratinocytes from either normal or psoriatic patients. Keratinocytes from 9 of 12 normal skin samples underwent double-stranded DNA fragmentation after being in methylcellulose for 18 to 24 hours, which contrasts with keratinocytes from lesions of psoriasis where only 1 of 13 of the skin samples had these changes. Third, two-color immunofluorescence staining of psoriatic plaques revealed that numerous TUNEL-positive keratinocytes were also positive for proliferating cell nuclear antigen and Ki-67 antigens and that by flow cytometry TUNEL-positive keratinocytes obtained from psoriatic plaques possessed a DNA content profile indicative of proliferating and not dying cells. These results demonstrate that keratinocytes within psoriatic plaques do not have double-stranded DNA breaks, that they have a prolonged capacity to resist induction of apoptosis compared with normal-skin-derived keratinocytes or cultured HaCat cells, and that caution is necessary for proper interpretation related to detection of 3'-OH DNA ends (ie, TUNEL positivity) in skin, as it can be associated with DNA synthesis as well as cell death.