Pressure ulcers are an important health care problem. Along with high morbidity and mortality, the costs associated with the care of a patient with pressure ulcers are significant. Only recently has clinical research focused on the nature and process of healing in these chronic wounds. This descriptive study examined factors associated with healing in a convenience sample of 106 nursing home patients with pressure ulcers. Guided by a framework that characterized chronic wound healing as a process frequently delayed or interrupted, the study examined the relationships among 12 selected patient, wound, and treatment factors and healing. Pearson r correlations of all independent variables and healing were calculated. Lower body temperature, higher weight, lower ulcer stage, and smaller amounts of wound exudate were significantly correlated to healing (p<.05). No treatment factors were significantly correlated to healing. Regression analyses resulted in four major findings. First, when patient variables were considered, lower body temperature and higher weight together predicted 9% of the variance in healing. Second, when wound variables were considered, 19% of the variability in healing was explained by the stage of the ulcer; specifically, lower stage (less severe) wounds demonstrated better healing than higher stage wounds. Third, when treatment variables were considered, only time on a pressure-relieving bed was significant, explaining 6% of the variance in healing. However, it was shorter time on the bed that was associated with improved healing due to the correlation of high stage wounds with bed therapy. Fourth, when all variables were considered, wound stage, patient weight, and mean body temperature together explained 25% of the variability in healing; specifically, this last finding indicated that lower wound stage, higher patient weight, and lower mean body temperature during the healing observation predicted improved healing. Stage was the single strongest predictor, explaining fully 19% of the variability in observed healing. Healing was not predicted by age, wound location, number of illnesses, use of tube feedings, use of pressure relieving beds, mean arterial pressure, or Braden scale score.