Abstract Introduction THN3 was the first reported parallel-arm, randomized controlled trial of hypoglossal nerve stimulation (HGNS), wherein targeted HGNS (THN) was shown to be safe and efficacious at ameliorating sleep disordered breathing in selected patients with moderate to severe OSA and BMI≤35kg/m2 without screening with drug-induced sleep endoscopy. A secondary THN3 objective was to identify baseline characteristics which enhance therapeutic response. Methods Predictors of long-term (Month 12/15) apnea-hypopnea index (AHI) response (reduction ≥50% to ≤20/hr) were identified in stages. Baseline characteristics, including polysomnography parameters, were screened one at a time using univariate logistic regression. Variables significant at α=0.25 were further selected using stepwise multivariate logistic regression. Candidate predictors were then optimized by manual, incremental threshold adjustment to maximize AHI response rate (RR). Optimized predictors were ranked according to the AHI RR and size of the corresponding subpopulation, with a goal of capturing at least 40% of the overall THN3 cohort. Results Baseline sleep efficiency (SE) in the normal range (>85%) achieved the highest AHI RR while retaining 42% (58/138) of the THN3 cohort. Similar to the overall study population, subgroup subjects were predominantly middle-aged (age 54±7 years), male (83%), Caucasian (86%) and overweight/obese (BMI 30±3 kg/m2). Short-term (Month 4) AHI RRs in the Treatment and Control groups were 59.5% and 20.0%, respectively, with long-term AHI RR at 50.9%. Median AHI was reduced from 35.1 to 16.0, as compared to 29.3 to 9.0 in the STAR trial of distal HGNS. ODI RRs (reduction ≥25%) for Treatment and Control were respectively 73.0% and 30.0% at Month 4 and 64.9% at Month12/15. Enhanced response carried through to larger point estimates for improvement in secondary variables, including the Epworth Sleepiness Scale and the Functional Outcomes of Sleep Questionnaire, exceeding minimum clinically important differences thresholds of 2-3 and 1.7-2.0, respectively. Conclusion Normal SE at Baseline was associated with enhanced AHI and ODI RR as well as patient-reported outcome measures for patients receiving THN therapy, providing results similar to those obtained with distal HGNS. Future studies, including the ongoing OSPREY trial, will be needed to prospectively validate these findings. Support (If Any) LivaNova