We conducted a prospective controlled study of the clinical and biological determinants of the mental status abnormalities in 139 patients with Guillain-Barré syndrome (GBS) and 55 patients without GBS placed in the intensive care unit (ICU controls). There were mental status changes in 31% of GBS patients and in 16% of controls (odds ratio = 2.3; P = 0.04). In GBS patients, they included vivid dreams (19% ), illusions (30% , including an illusory body tilt), hallucinations (60% , mainly visual) and delusions (70% , mostly paranoid). They appeared a median 9 days after disease onset (range 1- 40 days, during the progression or the plateau of the disease), and lasted a median 8 days. Seven (16% ) patients experienced the symptoms before their admission to the ICU. Hallucinations were frequently hypnagogic, occurring as soon as the patients closed their eyes. Autonomic dysfunction, assisted ventilation and high CSF protein levels were significant risk factors for abnormal mental status in GBS patients. CSF hypocretin- 1 (a hypothalamic neuropeptide deficient in narcolepsy) levels, measured in 20 patients, were lower in GBS patients with hallucinations (555 ± 132 pg/ml) than in those without (664 ± 71 pg/ml, P = 0.03). Since the mental status abnormalities had dream-like aspects, we examined their association with rapid eye movement sleep (REM sleep) using continuous sleep monitoring in 13 GBS patients with (n = 7) and without (n = 6) hallucinations and 6 tetraplegic ICU controls without hallucinations. Although sleep was short and fragmented in all groups, REM sleep latency was shorter in GBS patients with hallucinations (56 ± 115 min) than in GBS patients without hallucinations (153 ± 130 min) and in controls (207 ± 179 min, P < 0.05). In addition, sleep structure was highly abnormal in hallucinators, with sleep onset in REM sleep periods (83% ), abnormal eye movements during non-REM sleep (57% ), high percentages of REM sleep without atonia (92 ± 22% ), REM sleep behaviour disorders and autonomic dysfunction (100% ), reminiscent of a status dissociatus. The sleep abnormalities, that were almost absent in non-hallucinated GBS patients, were not exclusively related to ICU conditions, since they also appeared out of ICU, and were reversible, disappearing when the mental status abnormalities vanished while the patients were still in ICU. In conclusion, the mental status abnormalities experienced by GBS patients are different from the ICU delirium, are strongly associated with autonomic dysfunction, severe forms of the disease and possibly with a transitory hypocretin-1 transmission decrease. Sleep studies suggest that mental status abnormalities are wakeful dreams caused by a sleep and dream-associated disorder (status dissociatus). We conducted a prospective controlled study of the clinical and biological determinants of the mental status abnormalities in 139 patients with Guillain-Barré syndrome (GBS) and 55 patients without GBS placed in the intensive care unit (ICU controls). There were mental status changes in 31 GB of patients and in 16% of controls (odds ratio = 2.3; P = 0.04). In GBS patients, they included vivid dreams (19%), illusions (30%, including an illusory body tilt), hallucinations %, mainly visual) and delusions (70%, mostly paranoid). They have a median 9 days after disease onset (range 1- 40 days, during the progression or the plateau of the disease), and lasted a median 8 days. Seven (16%) patients experienced the symptoms before their admission to the ICU. Hallucinations were frequently hypnagogic, occurring as soon as the patients closed their eyes. Autonomic dysfunction, assisted ventilation and high CSF protein levels were significant risk factors for abnormal mental status in GBS patients. CSF hypocretin- 1 (a hypothalamic neuropeptide deficient in narcolepsy) levels, measured in 20 patients, were lower in GBS patients with hallucinations (555 ± 132 pg / ml) than in those without (664 ± 71 pg / ml, P = 0.03). Since the mental status abnormalities had dream-like aspects, we examined their association with rapid eye movement sleep (REM sleep) using continuous sleep monitoring in 13 GBS patients with (n = 7) and without (n = 6) Hallucinations and 6 tetraplegic ICU controls without hallucinations. Although sleep was short and fragmented in all groups, REM sleep latency was shorter in GBS patients with hallucinations (56 ± 115 min) than in GBS patients without hallucinations (153 ± 130 min) and in controls (207 ± 179 min, P <0.05). In addition, sleep structure was highly abnormal in hallucinators, with sleep onset in REM sleep periods (83%), abnormal eye movements during non-REM sleep REM sleep without atonia (92 ± 22%), REM s leep behaviour disorders and autonomic dysfunction (100%), reminiscent of a status dissociatus. The sleep abnormalities, that were almost absent in non-hallucinated GBS patients, were not exclusively related to ICU conditions, since they also occurred out of ICU, and were reversible , disappearing when the mental status abnormalities vanished while the patients were still in the ICU. In conclusion, the mental status abnormalities vanished while the patients were still in the ICU. In conclusion, the mental status abnormalities by GBS patients are different from the ICU delirium, are strongly associated with autonomic dysfunction, severe forms of the disease and possibly with a transitory hypocretin-1 transmission decrease. Sleep studies suggest that mental status abnormalities are wakeful dreams caused by a sleep and dream-associated disorder (status dissociatus).