INTRODUCTION: Orthodontic patients can experience pain and discomfort on the oral mucosa from trauma caused by friction with the brackets and the wires. In this split-mouth design, single-blind randomized controlled trial, we aimed to investigate whether brackets with a self-snapping customized plastic shield would induce less mucosa alteration and discomfort than those without the shield. METHODS: The overall sample comprised 42 patients (22 female, 20 male) from a government-funded orthodontic practice, with a mean age of 16.7 years. Eligibility criteria included, among others, no history of mouth ulcers or systemic diseases. Customized shields for the maxillary and mandibular brackets were fabricated and inserted on one side of the mouth. The null hypothesis was that bracket shielding would have no advantage. The primary outcomes were mucosal and discomfort assessments. As the secondary outcome, the numbers of spontaneous detachments of the shields were reported. Treatment allocation was mainly implemented using a random number table for selection of the intervention side. Only the raters in charge of assessing the oral mucosa were blinded to the side of the mouth where the shields had been placed. The mucosa was assessed by 3 calibrated raters at the following time points: immediately before bracket placement (baseline assessment, T0), 3 days after delivering the shields (direct assessment of intervention, T1), and 4 days after removal of the shields (indirect assessment of intervention, T2). The raters used a newly devised yardstick in which the higher the score, the more severe the alteration. Discomfort was assessed at T1 and T2 using a visual analog scale. The Mann-Whitney U test was performed at the 5% level of significance. RESULTS: Of 60 patients, 42 were eligible, and 35 were randomly selected to have one side of the mouth receive the intervention. Two patients discontinued the intervention at T1, and 5 stopped at T2. Seven additional patients were recruited and completed all time points. Thus, 42 patients participated at T0, 40 at T1, and 35 at T2. Thirty-five patients participated at all time points. At T1, no statistically significant difference in terms of mucosa alteration was observed between the 2 sides (median of all differences [MD], 0.0; 95% CI, 0.0-1.0; P = 0.11). The same occurred at T2 (MD, 0.0; 95% CI, 0.0-0.0; P = 1.00). The comfort level was statistically higher at T1 on the shielded side (MD, 14.0; 95% CI, 1.0-36.0; P = 0.04), whereas no difference was observed at T2 (MD, 0.0; 95% CI, 0.0-1.0, P = 0.81). No serious harm was observed. CONCLUSIONS: The customized bracket shields were effective in reducing discomfort during the first 3 days of orthodontic treatment despite no significant difference in terms of visible mucosa alteration. REGISTRATION: This trial was not registered. PROTOCOL: The protocol was not published before trial commencement. FUNDING: Expenses for the fabrication of the shields were covered by the main author (L.P.B.P.). Orthodontic materials were from the Center for Dental Specialties in Cajazeiras, Brazil.