Aims: To determine whether stretch duration and applied force affect maximal incisal opening (MIO) in patients with trismus secondary to head and neck cancer treatment. Methodology: Prospective, multi-centre adaptive factorial randomised controlled trial. Participants were randomised to one of six intervention arms applying calibrated loading (three stretch durations: 30, 60, 120 seconds; two force levels: 30 N or 60 N[SP1.1][EB1.2]) using Restorabite® versus the control arm applying uncalibrated loading for 30 seconds with stacked tongue depressors for 10 weeks. The primary outcome measure was change in MIO from baseline to 10 weeks. Analysis of covariance adjusted for baseline MIO and treatment modality. Results: Of 328 adults with trismus (MIO < 35 mm) who were recruited across seven Australian centres, 280 participants completed the intervention. Mean baseline MIO was 23.1 mm. Greater improvement in MIO was observed in the intervention arms using calibrated loading compared to the control (10.5 mm vs 6.9 mm; adjusted difference (β) = 3.5 mm; 95% CI 1.3–5.8; p = 0.002). The largest effect was observed in the high-force, long-duration arm, where five 120 second stretches, using a force of 60 N, were completed each day (β = 5.3 mm; 95% CI 2.4–8.1; p < 0.001). Both higher force (p = 0.001) and longer duration (p = 0.001) were independently associated with greater gains. Single-modality cancer treatment (surgery alone or (chemo)radiotherapy alone) was also associated with larger improvement (β = 2.4 mm; p = 0.002). No serious adverse device-related events occurred. Conclusion: Calibrated loading applying high force and longer duration produces superior mouth-opening gains compared with uncalibrated loading. This study is the first high quality evidence that trismus rehabilitation follows a dose–response relationship, establishing a foundation for personalised mechanotherapy.