A nicotine-impaired spinal fusion rabbit model. To examine whether controlled delivery of morselized absorbable collagen sponge recombinant human bone morphogenetic protein-2 (rhBMP2) in a delayed manner postsurgery would allow for improved bone healing. The current delivery method of rhBMP-2 during surgery causes a burst of rhBMP-2, which is not sustained. Given that bone morphogenetic protein 2 (BMP-2) expression peaks later in the fusion process, there may be the benefit of delivery of rhBMP-2 later in the healing process. Sixteen male 1-year-old rabbits underwent a posterolateral spinal fusion with iliac crest bone graft at L5-L6 while being given nicotine to prevent spinal fusion as previously published. Eight were controls, whereas 8 had morselized rhBMP-2 (4.2 mg) injected at the fusion site at 4 weeks postoperatively. Histologic, radiologic, and palpation examinations were performed at 12 weeks to determine fusion status and the volume of bone formed. Hematoxylin and eosin stains were used for histology. A Student t test was used to compare the computed tomography scan measured volume of bone created between the control cohort (CC) and rhBMP-2 delayed delivery cohort (BMP-DDC). Of the total, 7/8 rabbits in the BMP-DDC and 5/8 rabbits in the CC formed definitive fusion with a positive palpation examination, bridging bone between transverse processes on computed tomography scan, and an x-ray showing fusion. Histologic analysis revealed newly remodeled bone within the BMP-DDC. There was an increased average volume of bone formed within the BMP-DDC versus the CC (22.6 ± 13.1 vs 11.1 ± 3.6 cm 3 , P = 0.04). Our study shows that injectable morselized absorbable collagen sponge/rhBMP-2 can create twice as much bone within a nicotine-impaired rabbit spine fusion model when delivered 4 weeks out from the time of surgery.