The sensitivity of several neurophysiological and cognitive tests to different levels of hypoxia was investigated. Cerebral hypoxia in healthy volunteers may be a disease model for dementia or other forms of brain dysfunction. Twelve healthy subjects were included in a randomized, single-blind, placebo-controlled, three-period cross-over trial. They received three air/N 2 gas mixtures via mask breathing [aimed at peripheral oxygen saturation (SPO 2 ) values of > 97% (placebo), 90% and 80%, with normal end-tidal CO 2 ]. Central nervous system effects were tested regularly for 130 min by saccadic and smooth pursuit eye movements, electro-encephalogram, visual analogue scales and cognitive tests. Treatments were well tolerated. Compared to SPO 2 90%, SPO 2 80% reduced saccadic peak velocity by 16.4 /s [confidence interval (CI) -26.3, -6.4], increased occipital delta power by 14.3% (CI 3.6, 25.1), and significantly increased most cognitive reaction times. SPO 2 80% also decreased correct responses for the binary choice task and serial word recognition [-1.3 (-2.2, -0.3) and -3.5 (-6.2, -0.8), respectively] compared to SPO 2 90%. Cognitive performance was decreased by SPO 2 80% and increased by SPO 2 90% compared to placebo. Sensitive effect measurements can be identified for these interventions. The applicability as a model for cognitive impairment should be investigated further.