The early assessment of new symptomatic drugs against Alzheimer's disease remains difficult because of the lack of a predictive end-point. The use of a battery including different parameters could improve this early development. In order to test the reverse effect of symptomatic drugs in healthy volunteers, scientists have developed new experimental paradigms to artificially induce transient cognitive impairments in healthy volunteers akin to those observed in Alzheimer's disease, i.e. Cognitive Challenge Models. In this context, transient hypoxia could be a relevant Cognitive Challenge Model. The deleterious effects of hypoxia on cognition, as described in the literature, should be considered carefully since they are usually assessed with different populations that do not have the same hypoxic sensitivity. Hypoxia can be obtained by the means of two different methods: normobaric and hypobaric hypoxia. In both designs, cognitive changes can be directly modulated by the severity of hypoxic levels. The purpose of this review is to gather existing support on the application of hypoxia within different cognitive domains and to highlight the scientific interests of such a model to predict and select promising drug candidates. We aimed at reviewing in detail the methods, designs and cognitive paradigms used in non-pharmacological hypoxia studies. Probing the four main cognitive functions will allow identifying the extent to which different hypoxia designs selectively compromise cognitive functioning. For each cognitive process, the convergent and divergent results are discussed in terms of paradigm differences whereas we will focus on defining the optimal methodology for obtaining the desired effects.