Sleepiness in patients with obstructive sleep apnea (OSA) is associated with accidental and economic burden, as well as cardiovascular risk. Despite OSA treatment, 10-28 % of patients report residual sleepiness. Its determinants, as well as those of objective impaired alertness remain poorly known. In this study, we investigated factors associated with residual subjective sleepiness and objective impaired alertness in patients treated for OSA. Consecutive OSA treated patients referred for maintenance of wakefulness tests (MWT) at a tertiary university center were recruited between 2017 and 2020. Clinical data and polysomnography parameters were compared between patients with vs without subjective sleepiness (Epworth Sleepiness Scale, ESS≥11) and those with vs without impaired alertness (at least one trial with sleep onset on MWT). A multivariate logistic model was used to assess explanatory variables of MWT and ESS results. We included 141 patients, of whom 12.8 % had both subjective sleepiness and objective impaired alertness, 17.7 % objective impaired alertness only and 9.2 % subjective sleepiness only. Self-reported history of car accident/near miss, smoking history and ESS≥11 were significantly associated with objective impaired alertness whereas residual Apnea-hypopnea Index and CPAP use were not. The only significant variable associated with ESS at the time of MWT evaluation was initial ESS. Patients with objective impaired alertness only were more often smokers (52 % vs 19 %, p = 0.01), had a higher body mass index (BMI) (32 vs 29 kg/m, p = 0.05), and showed lower initial ESS (11 vs 13, p < 0.01). More than one third of OSA treated patients referred for MWT have objective impaired alertness and/or subjective sleepiness. Our findings highlight the need for a comprehensive medical assessment including accident history, subjective sleepiness and comorbidities. Particular attention should be paid to smoking patients with high BMI, who are at risk of impaired alertness with no report of subjective sleepiness.< Réduire