One hundred, twenty-four patients undergoing lumbar or lumbosacral fusion for degenerative conditions were entered into a prospective study. The patients were randomly assigned to one of three treatment groups. Group I underwent posterolateral fusion using autogenous bone graft. Group II had autogenous posterolateral fusions supplemented by a semi-rigid pedicle screw/plate fixation system (Luque II; Danek Medical, Memphis, Tennessee). Group III patients underwent posterolateral autogenous fusion with a rigid pedicle screw/rod fixation system (Texas Scottish Rite Hospital [TSRH]-Danek Medical, Memphis, Tennessee). All the patients were operated on by the same surgeon, identical bone grafting technique was used in all, and all were treated in an identical fashion postoperatively. Fusion status was determined from the anteroposterior, oblique, and flexion-extension radiographs obtained at 1 year. Clinical results were rated as excellent if the patients were pain-free and had returned to work; good if the patients had mild backache requiring non-narcotic analgesics and had returned to work; fair if continuing back pain prevented a return to work; or poor if the pain was worse than that which the patient experienced preoperatively or the patient required revision surgery. Nine patients who were originally assigned to Group II or Group III were placed in Group I intraoperatively. This change was due to the identification of severe osteopenia and the determination that pedicle screw purchase was poor. One patient was lost to follow-up. Thus, 51 patients were in Group I, 35 in Group II, and 37 in Group III. Follow-up ranged from 9 to 28 months, averaging 16 months.(ABSTRACT TRUNCATED AT 250 WORDS)