This retrospective study of 48 spinal cord-injured patients with pressure ulcers seen at a tertiary referral Veterans Hospital spinal cord injury unit between 1992 and 1997 correlates a number of variables (co-morbid conditions, nutritional status, smoking history, type of repair performed, type of bed used postoperatively, ulcer location and severity, duration of postoperative antibiotic therapy, time elapsed before sitting rehabilitation began, and length of hospital stay) with ulcer repair outcome measures, including postoperative systemic and wound-healing complications, recurrence rates, and the development of new ulcers at different sites. Surgical complication rates were high, occurring in 19 patients (39.6%), and ulcer recurrence or new ulcer development occurred in 38 patients (79.2%). Correlations were found between ulcer location and postoperative wound separation and the length of hospitalization. The hospital course was shorter if the ulcer was new rather than recurrent. Other than the finding that chronic smokers had longer courses of antibiotic therapy, smoking did not correlate statistically with other outcome variables, including wound-healing complications. No significant correlations were found between any postoperative systemic or wound complications, ulcer recurrence, or new ulcer development and patient age, level of spinal cord injury, number of ulcers and grade, laboratory values, mental status, cardiac or pulmonary disease, diabetes, and presence or absence of osteomyelitis.