Background Functional restoration following resection or traumatic injury to the mandible depends on the reliability of the bony reconstruction to heal primarily and support endosseous implants. Although vascularized bone flaps (VBF) and nonvascularized bone grafts (NVBG) are both widely used to reconstruct the mandible, indications for each remain ill-defined. The purpose of this study was to compare bone graft/flap healing and success of implant placement in patients reconstructed with VBF versus NVBG. Methods Over the past 10 years, 75 consecutive mandibular reconstructions were performed (26 free bone grafts, 49 vascularized bone flaps). Etiology of the defect, history of irradiation, bone defect size, number of operations, graft/flap success, and dental implant success rates were determined and compared. Bone graft/flap success was defined as complete bony union. Implant success was defined as complete osseointegration. Mean follow-up was 3 years. Results Free flaps were used primarily for malignant disease (78%, 38/49). Bone grafts were used primarily for benign disease (88%, 23/26). History of prior irradiation: 11% (3/26) NVBG versus 45% (22/49) VBF. Length of bony defect (mean): 8.1 cm NVBG versus 9.4 cm VBF. Successful bony union, any size defect: 69% (18/26) NVBG versus 96% (47/49) VBF (p < .0005); lateral defects only: 75% (15/20) NVBG versus 100% (17/17) VBF (p < .05). Number of operations to achieve bony union (mean), any size defect: 2.3 NVBG versus 1.1 VBF (p < .001); lateral defects only: 1.9 NVBG versus 1.0 VBF (p < .005). Twenty-two patients (29%) had a total of 104 endosseous implants placed (NVBG: 8 patients, 33 implants; VBF: 14 patients, 71 implants). Immediate implants placed: 0/33 NVBG versus 54% (38/71) VBF. Overall implant success: 82% (27/33) NVBG versus 99% (70/71) VBF (p < .0001). Implant success in VBF patients with a history of RT: 100% (15/15). Conclusions Despite the fact that patients reconstructed with VBFs were older, had larger defects, and were treated primarily for malignant disease and therefore had a higher incidence of irradiation to the affected mandible than in patients treated with NVBGs, the incidence of bony union was higher, requiring fewer operations to achieve union, and the implant success rate was significantly greater than for NVBG patients. Results were similar when considering lateral defects only. Based on these results, VBFs are indicated in most cases of mandibular reconstruction; NVBGs are effective for short bone defects (<5–6 cm), in nonirradiated tissue, and/or in patients determined to be too medically compromised to tolerate the additional operative time required for a free-flap reconstruction. © 1999 John Wiley & Sons, Inc. Head Neck 21: 66–71, 1999.