Treatment of spinal fractures causing paralysis in patients with ankylosing spondylitis is difficult. A multicenter concurrent study evaluated 59 patients--22 managed operatively and 37 nonoperatively. The two groups were compared for neurologic outcome, complications, mortality, and length of stay. The results indicated that patients in the nonoperative group had a significantly shorter length of stay and, therefore, a significantly lower cost of care. No other differences between operative and nonoperative groups were identified in regard to other outcome variables. Results of descriptive analyses of patient characteristics and treatment choices have significant implications for practitioners.