Two groups of patients with diseased lumbar spines treated by the Steffee variable screw placement system were studied, in order to compare concomitant posterolateral fusion and posterior lumbar interbody fusion. The use of posterolateral fusion resulted in less invasiveness than posterior lumbar interbody fusion, with regard to operating time and blood loss. With the posterolateral fusion method, correction of olisthesis or kyphotic deformity, if attained at all, was difficult to maintain. Posterior lumbar interbody fusion produced better correction and maintenance, even in osteoporotic cases. Bone union was achieved at an average of 11 months in only 60% of the posterolateral fusion group, whereas it was complete within a significantly shorter period (average; 6 months) in 91% of the posterior lumbar interbody fusion group. Breakage or loosening of screws occurred in 14% of the posterolateral fusion group. Deep wound infection and adhesive arachnoiditis, which were never seen in the posterolateral fusion group, developed in 3% of the posterior lumbar interbody fusion group. Posterior lumbar interbody fusion is considered to be a better method of bone grafting than posterolateral fusion, when used with the Steffee variable screw placement system.