A randomized clinical trial was conducted to evaluate the efficacy of three commonly employed forms of traction in the treatment of cervical spine disorders. One hundred consenting men and women with disorders of the cervical spine were randomly assigned to one of four treatment groups, static traction, intermittent traction, manual traction, or no traction. All patients, regardless of group assignment, were seen twice weekly. The four groups were shown to be similar with regard to age, sex, diagnosis, chronicity, and prescores on the seven outcome measures. Although the entire cohort of neck patients, regardless of group assignment, improved significantly on all the outcome variables over the 6-week period, patients receiving intermittent traction performed significantly better than those assigned to the no traction group in terms of pain (P = 0.03), forward flexion (P = 0.01), right rotation (P = 0.004) and left rotation (P = 0.05).