Objective: This study prospectively evaluates the effect on sternal vascularity of harvesting the left internal thoracic artery. Methods: Twenty- four consecutive patients undergoing primary coronary artery bypass grafting were studied. One patient's procedure was altered during the operation, and he was eliminated from the study. The patients were prospectively randomized to receive a skeletonized internal thoracic artery (group I, n = 11) or a pedicled internal thoracic artery (group II, n = 12) graft. Each patient underwent a preoperative technetium 99 methylene diphosphonate bone scan using single photon emission computed tomography. The ratio of the mean counts per pixel on the left side of the sternum was compared with the mean counts per pixel on the right side. Postoperatively, all patients had a second scan, and sternal uptake was compared with the preoperative uptake. Results: No significant differences in preoperative and operative variables were observed between the groups. A statistically significant reduction in blood flow to the left side of the sternum was shown postoperatively in group II compared with group I (0.61 ± 0.11 vs 0.85 ± 0.09; P < .001). Multivariable logistic regression analysis of preoperative and operative variables revealed only a pedicled left internal thoracic artery to be associated with a 20% or more reduction in left-to-right sternal activity ratio (odds ratio, 100; 70% confidence limits, 22-465; P = .002). Conclusion: A pedicled left internal thoracic artery graft to the left anterior descending artery reduces blood flow to the left side of the sternum during the acute postoperative period. This does not occur when the left internal thoracic artery is skeletonized.