In August, 2004, a 71-year-old man was admitted to our hospital, with a hoarse voice and “noisy sleep”. His medical history was essentially uneventful: mild hypertension, for which he took valsartan; and a brief episode of laryngitis in the early 1980s. He had dysphonia, and persistent nocturnal inspiratory stridor, but no diurnal dyspnoea or stridor. Neurological examination revealed slight diffi culty with tandem gait, which we considered normal, given the patient’s age. Blood tests, CT of the chest, and MRI of the brain and cervical spine showed nothing of note. Laryngoscopy revealed bilateral vocalfold motion impairment: we diagnosed idiopathic bilateral recurrent-laryngeal-nerve palsy. The patient was discharged 8 days after admission. 4 months later, the patient was admitted again, with unsteadiness. His stridor was less severe than before. On detailed questioning, the patient reported erectile dysfunction, urinary urgency, and a recurring nightmare of being chased by a bear, during which he had hit his wife. Examination showed dysphonia and cerebellar dysfunction (dysdiadochokinesia and ataxia). We found retropulsion, consistent with Parkinsonism, but no tremor, rigidity, or bradykinesia. We noted asymptomatic postural hypotension (systolic blood pressure dropped from 130 mm Hg to 98 mm Hg). We found no other abnormalities, other than brisk knee jerks. Lumbar puncture, and electromyography of the anal sphincter, showed no abnormality. We diagnosed multiple system atrophy. Videopolysomnography, with sound recording, revealed prominent inspiratory stridor and obstructive sleep apnoea (34 events per h, causing oxygen saturations to decrease transiently to 75%). During REM sleep, tonic muscle activity remained, and the patient spoke and pointed with his right arm, confi rming that he could move during his nightmares: he had REM sleep behaviour disorder. We prescribed clonazepam (250 μg) for the sleep disorder, and treated sleep apnoea with bilevel positive airway pressure, which was shown by polysomnography to be highly eff ective. However, as expected, the patient’s underlying condition has continued to worsen. When he was last seen, in April, 2007, his ataxia had worsened; he fell often when trying to walk unaided. His postural hypotension had become symptomatic. Nonetheless, he slept well, and was not sleepy in the daytime. He had not dreamt of the bear for several months. Multiple system atrophy is a term used to describe syndromes formerly known as Shy-Drager syndrome, striatonigral degeneration, and olivoponto cerebellar atrophy. It is a neurodegenerative disease—which, like Parkinson’s disease and Lewy-body dementia, is characterised histopathologically by cytoplasmic accumu la tion of α-synuclein. Patients with multiple system atrophy have motor and autonomic dysfunction. In 80% of patients, motor features are mainly parkinsonian; other patients have cerebellar dysfunction. Electro myography of the anal sphincter often gives abnormal results, but is not required for diagnosis. Manifestations of autonomic dysfunction include postural hypotension, urinary urgency, and erectile dysfunction. REM sleep behaviour disorder has been found in over half of patients with multiple system atrophy. Vocal-fold motion impairment develops in around 30% of patients; it was thought to be caused by impaired abduction of the vocal cords, attributable to degeneration of the nucleus ambiguus and recurrent laryn geal nerve. However, such a mechanism is diffi cult to reconcile with the spontaneous improvement in our patient’s stridor, and with the worsening of symptoms at night. Recent work supports a diff erent mech an ism: dystonia of the laryngeal muscles during sleep, ie, inappropriate adduction of the cords. Since ventilatory support was introduced as a treatment, tracheostomy is rarely required.