Background New generation positional therapy devices provide vibrotactile feedback to patients with positional obstructive sleep apnoea (POSA), reducing supine sleep time and sleep apnoea severity. Longer-term effects on POSA severity, sleepiness and quality of life (QoL) are unclear. Methods A randomised, parallel, double-blinded trial compared neck-worn positional therapy with sham-positional therapy over 3 months (ClinicalTrials.gov: NCT04153240). Adult patients with POSA (apnoea/hypopnoea index (AHI) >5 events/hour, 2:1 when supine vs non-supine) were randomised (1:1). The primary endpoint was AHI at 3 months, positional versus sham. Secondary analyses: interaction between the treatment effect and age; QoL, including Epworth Sleepiness Scale (ESS) and Pittsburgh Sleep Quality Index (PSQI). Results Between October 2019 and August 2022, 120 patients with median baseline AHI of 12.8 events/hour (IQR 9.2–18.5) were randomised; 59 to positional therapy and 61 to sham; 92 (77%) completed the trial. Positional therapy significantly reduced the AHI by −4.41 events/hour (95% CI −7.77 to –1.06; p=0.011) compared with sham, a 34% improvement. There was a significant improvement in PSQI: −1.0 (95% CI −2.1 to 0.0; p=0.04), but not ESS: −0.6 (95% CI −1.8 to 0.6; p=0.3), with positional therapy compared with sham (baseline ESS 8.8). Similar results were seen in younger (18–64) and older (≥65) age groups. Patients’ bed partners reported improvements in snoring and sleep quality for the patient and themselves. Over half of participants using the active device opted to continue. Conclusion Neck-worn positional therapy reduced the severity of OSA and improved sleep quality but not sleepiness, over 3 months. Bed partner’s reported improvements in snoring and sleep quality. Trial registration number NCT04153240.