Pressure ulcers appear as localized chronic wounds in the middle of normal functioning skin. This study focuses on pressure ulcer fibroblasts cultured from the ulcer bed, ulcer margin, and normal, nonulcerated skin adjacent to the wound. From these areas we show that pressure ulcer fibroblasts become prematurely senescent. Verification of senescence was based on failure of cell populations to undergo a 0.5 population doubling after 1 week in culture, light microscopic appearance of late-passage senescent cells in culture, light microscopic verification of senescence in vivo and in vitro by anti-terminin staining, and ultrastructural identification of senescent fibroblasts by anti-terminin colloidal gold labeling. Although not all ulcer fibroblasts are senescent, there is a range of proliferative ability. The senescent phenotype of ulcer fibroblasts remains intact in vitro as these cells remain viable but are unable to complete DNA synthesis. Fibroblast senescence is not uniform in chronic wounds but varies within areas of the ulcer bed and from patient to patient. Although ulcer fibroblasts exhibit limited proliferative ability, fibroblasts from the ulcer margin and adjacent normal skin show a continued ability to divide. However, in all areas of the wound, the mean generation time of the fibroblast population is longer than commonly observed for cultured human diploid fibroblasts. At first passage, the mean generation time for fibroblasts from all patients is significantly different (p = .001, analysis of variance) among fibroblasts from the ulcer bed (4.2 2.2 days), ulcer margin (2.9 +/- 1.3 days), and adjacent normal skin (1.9 +/- 0.6 days). Analysis of the three cell groups by the post hoc Tukey test, using corrected p values shows that the difference between mean generation times among fibroblast populations from adjacent normal skin and ulcer bed are significantly different (p < 0.05), whereas the difference between fibroblasts from adjacent normal skin and ulcer margin and ulcer margin and ulcer bed are not significantly different (p > 0.05). These data may help to explain the poor response of certain pressure sores to aggressive medical treatment.