Since the initial description by Michelet et al. and research by Champy et al. the placement of a single, four-hole, monocortical, osteosynthesis plate has been considered an acceptable method of fixation for a fracture of the mandibular angle. We investigated the null hypothesis that there is no difference in the incidence of postoperative removal of an infected plate between miniplates placed on the mandibular external oblique ridge and those placed on the buccal surface of the mandible through a transbuccal approach to treat a fracture of the angle of the mandible. Patients were randomised to having their angle fractures treated with a ridge plate placed intraorally or transbucally. Other variables were investigated including the effect of smoking, drinking alcohol, oral hygiene, and the method of holding the reduction on removal of the plate, occlusal outcome, and degree of preoperative anatomical displacement and postoperative reduction. We also studied the operating time required for the two techniques, the effect of the presence and consequent removal of a wisdom tooth in the line of the fracture, and the effect of delay in taking the patient to theatre for subsequent removal of the plate for infection. Of the 261 angle fractures 34 (13%) plates were removed because of infection, and 6 of these (18%) required a further period of fixation, such as intermaxillary fixation, to treat non-union. The transbuccal plate had a significantly lower postoperative infection rate (6/124, 5%) than the ridge plate (28/137, 20%) (p = 0.001). Smoking adversely affected the healing of angle fractures (p = 0.000). Displacement of fractures is related to the infection rate (p = 0.003), and there are no significant relations between delay in going to theatre or the presence and potential removal of a wisdom tooth in the line of the fracture and infection rate. There was a highly significant difference between the rate of removal of plates placed intraorally on the external oblique ridge, and plates placed transbucally (p = 0.000). Transbuccal plates were far less likely to need removal for infection than ridge plates, odds ratio 5.05.