Purpose: Considering the favourable prognosis of thyroid tumours, the importance of quality of life, including functional and voice outcomes, is critically important. The impact of remote-access thyroidectomy on postoperative voice function remains unclear, and direct comparative data across major approaches are limited. This study compared patient-reported and objective voice outcomes following robotic transoral (TO), robotic retroauricular (RA), and open hemithyroidectomy. Methods: This retrospective observational cohort study included patients undergoing hemithyroidectomy between January 2022 and August 2025. Voice outcomes were evaluated preoperatively and at 3 months postoperatively using the Voice Handicap Index-10 and acoustic analysis. Between-group comparisons and multivariable general linear modelling were performed to identify factors associated with postoperative voice outcomes. Results: A total of 150 patients were analysed (RA: 58; TO: 37; open: 55). Baseline VHI-10 and acoustic parameters were comparable across groups. The RA group demonstrated significantly higher postoperative VHI-10 scores and greater VHI-10 change scores compared with TO and open approaches ( p < 0.01). No significant differences were observed in postoperative acoustic parameters, including jitter, shimmer, noise-harmonic ratio, soft phonation index, pitch range, or fundamental frequency. Female subgroup analysis demonstrated similar findings. On multivariable analysis, postoperative VHI-10 remained to be only affected by the type of surgical approach. Conclusion: All three surgical approaches showed comparable objective acoustic voice parameters. However, TO and open hemithyroidectomy were associated with better patient-perceived voice outcomes at 3 months compared with RA. When clinical conditions are comparable, TO may be considered as a viable robotic approach in patients who prioritise short-term subjective voice outcomes.