. NASAL continuous positive airway pressure (nCPAP) is the ®rst-line therapy for obstructive sleep apnea syndrome (OSAS), a highly prevalent disorder associated with signi®cant morbidity and mortality rate (1). Patients with OSAS have in ammation of the upper respiratory tract (2), and nasal obstruction is commonly reported as a potential risk factor for the disease. In fact, patients with allergic rhinitis have a 1.8 higher risk of OSAS (3), and patients with seasonal allergic rhinitis have more sleep apnea during the pollen season (4). On the other hand, nasal blocking is one of the most frequent side-effects in the initial phase of nCPAP therapy (5). We report a case of late-onset rhinitis after nCPAP therapy for OSAS. A 56-year-old woman, treated for secondary hypothyroidism (after partial thyroidectomy due to multinodular goiter), was evaluated for suspected OSAS. She had marked daytime somnolence, with an Epworth score of 22 (normal less than 10, maximum 24). Her physical examination was unremarkable; she presented normal lung function but had resting hypoxemia (PaO2 66.4 mmHg). Routine blood analyses were normal, and thyroid function was evaluated ever 5±6 months, remaining within normal range throughout our observations. Diagnostic polysomnography showed marked sleep fragmentation due to frequent bouts of obstructive apnea and hypopnea (apnea/hypopnea index: 50), establishing the diagnosis of severe OSAS. Home nCPAP (Aria, Respironics, Monroeville, PA, USA) was then started at 8 cmH2O. She initially had good compliance with treatment, with no sideeffects and signi®cant symptomatic improvement was noted (Epworth score of 8 and PaO2 of 86 mmHg). After 6 months of treatment (October 1998), she began to have severe nasal symptoms: mucous rhinorrhea, sneezing, frontal headaches, and nasal obstruction. These symptoms were related to her use of the nasal mask and improved with the voluntary interruption of nCPAP. This noncompliance made OSAS symptoms recur (Epworth score of 20). There was no previous history of rhinitis or sinusitis, and no signs of infection were detected. Skin prick tests to common allergens and latex (four different commercial extracts) were negative. Total serum IgE, speci®c latex IgE, and an upper airway CT scan were normal. Nasal peak inspiratory ow (nPIF) (In-Check, Clement Clarke, Harlow, UK) was evaluated seven times daily before and after trying to restart nCPAP. The minimum daily value fell from 180 to 70 l/s, and nasal symptoms reappeared (rhinorrhea and local pain). Due to worsening OSAS symptoms, she agreed to another nCPAP ``challenge'', monitoring of the nasal symptoms score (NSS) and nPIF, analysis of nasal lavage, and active anterior rhinomanometry (Hortmann, Rhinomodul, Germany) before and after 7 days of use. Compliance with treatment (i.e., number of hours of effective treatment) was automatically recorded by Aria CPAP. These tests showed an increase of the symptoms score from 0 to 8.7, a fall in nPIF, and a decrease in nasal ow on nCPAP therapy (total expiratory air ow fell from 1059 to 769 ccm/s). Absence of in ammatory cells in nasal lavage was observed on both occasions. Several therapeutic approaches were unsuccessful; these included changing the nasal mask and several medications (oral antihistamines, local and oral vasoconstrictors, nasal ipatropium bromide, and topical steroids). The in uence of the air temperature was thought to be a possible cause of the symptoms, and ``natural warming'' of the nCPAP circuit was tried by placing it under bedclothes. This simple measure achieved complete resolution of nasal complaints (mean NSS: 1.3), with stabilization of nPIF recordings around 110 l/s, and a gain in nCPAP compliance (average hours of use: 6.3) with improved OSAS symptoms (Epworth score 3). Our case shows that a detailed investigation of nasal symptoms during nCPAP can eliminate this side-effect and improve compliance with nCPAP in OSAS. Serial nPIF monitoring is a useful tool for the objective evaluation of nasal complaints in patients undergoing nCPAP. Although the mechanisms involved in the nonallergic rhinitis re ex to nCPAP in this patient are not fully elucidated, the bene®cial effect of air warming suggests that the temperature of the inspired air has a role. This simple, inexpensive approach challenges recent reports recommending the use of heated humidi®ers (a costly solution) to resolve nasal symptoms of patients undergoing CPAP (6).