In 24 rabbits, the authors transplanted autologous cancellous bone to the anterior chamber of the eye. Half of the rabbits received nicotine and half received placebo (albumin) from mini-osmotic pumps that were implanted subcutaneously. Revascularization of the bone graft was evaluated postoperatively using ophthalmology slit-lamp and fluorescein anglography, and after sacrifice using microvascular silicone injection and histology. The hypothesis that nicotine inhibits the revascularization of bone graft because of its pharmacologic action on the microvasculature was tested. Pseudoarthrosis after spinal fusion occurs more frequently in smokers as compared with nonsmokers. Observations of the bone graft were made regarding the time after implantation when vessels within the graft were noted and the pattern of these vessels. Revascularization of the graft was graded based on the observed percent area of fluorescence after injection of fluorescein. Serum levels of nicotine were measured weekly. Colored silicone was injected at sacrifice to fix the vasculature of the bone graft. Histologic analysis of undecalcified sections was performed. Nicotine, as compared with placebo was associated with delayed revascularization within the graft, a smaller percent area of revascularization, and a larger number of grafts showing necrosis. Nicotine inhibits, but does not prevent, the revascularization of cancellous bone grafts. Inhibition of early revascularization by nicotine is proposed as the pathophysiologic mechanism by which smoking may adversely affect the healing of spinal fusions.