The use of allografts continues today for the purpose of spinal fusion. The literature is conflicting on the effectiveness of allografts in spinal fusion as compared with autografts. Numerous variables affect fusion, including age, sex, smoking status, type of surgery, and anterior versus posterior surgery. Therefore, we have undertaken a prospective study in which comparison was made between allograft and autograft in the same individual. The purpose of this prospective study was to compare autografts, frozen allografts, freeze-dried allografts, and a mixture of allograft and autograft in the same patient undergoing an instrumented posterolateral lumbar spine fusion. Twenty patients (nine men and 11 women with ages ranging from 29-72 years and a mean age of 43.5 years) underwent posterolateral fusions of the lumbar spine with pedicle screw instrumentation. An autogenous posterior iliac crest bone graft was placed on one side in each patient and an allograft on the other side. Assessing bone fusion quality from grades 1 to 4, the autograft side had a grade 1 solid fusion in 16 of 20 cases (80%). On the other hand, freeze-dried grafting resulted in grade 4 resorption in all seven cases. Frozen allografts resorbed in three of five cases, and partial fusions were achieved in the remaining cases. When a mixture of autograft and freeze-dried allograft was used, grade 1 solid fusion was achieved in four of eight cases and partial fusions were achieved in the others. Bone densitometry results also showed that autograft sites gave significantly greater bone density, followed by mixture, frozen allografts, and freeze-dried allografts in this order.(ABSTRACT TRUNCATED AT 250 WORDS)