Research on the HIV-related neuropsychological impairment has been marked by the variety and discrepancy of results, mainly in regards to asymptomatic phases. These discrepancies have been associated with the different types of methodology used in these studies. In this article, we review the main methodological problems present in studies about HIV-related neuropsychological impairment. Our aim is to gather the suggestions contained in the literature, to help overcome these problems in future studies in the field. These suggestions consider the samples used, the way in which they are selected, their size, the sampling methods used, the risk group studied, the criteria for inclusion/exclusion applied, or the selection of the control group. References are also included for the measuring instruments or neuropsychological tests used, statistical methods and criteria for defining the presence of neuropsychological impairment, and for dealing with confounding factors. There can be no doubt that a precise and thorough examination of these issues will make it possible to achieve more conclusive results on the incidence and nature of HIV-related neuropsychological impairment, and the factors that are associated with it.