Vascular cognitive impairment (VCI) includes vascular dementia (VaD), vascular mild cognitive impairment (VaMCI) and mixed dementia. In clinical practice, VCI concerns patients referred for clinical stroke or cognitive complaint. To improve the characterization of VCI and to refine its diagnostic criteria, an international group has elaborated a new standardized evaluation battery of clinical, cognitive, behavioral and neuroradiological data which now constitutes the reference battery. The adaption of the battery for French-speaking subjects is reported as well as preliminary results of the on-going validation study of the GRECOG-VASC group [Clinical Trial NCT01339195]. The diagnostic accuracy of various screening tests is reviewed and showed an overall sub-optimal sensitivity (<0.8). Thus, the general recommendation is to perform systematically a comprehensive assessment in stroke patients at risk of VCI. Furthermore,the use of a structured interview has been shown to increase the detection of dementia. In addition to the well known NINDS-AIREN criteria of VaD, criteria of VCI have been recently proposed which are based on the demonstration of a cognitive disorder by neuropsychological testing and either history of clinical stroke or presence of vascular lesion by neuroimaging suggestive of a link between cognitive impairment and vascular disease. A memory deficit is no longer required for the diagnosis of VaD as it is based on the cognitive decline concerning two or more domains that affect activities of daily living. Both VaMCI and VaD are classified as probable or possible. These new criteria have yet to be validated. Considerable uncertainties remain regarding the determinant of VCI, and especially the lesion amount inducing VCI and VaD. The interaction between lesion amount and its location is currently re-examined using recent techniques for the analysis of MRI data. The high frequency of associated Alzheimer pathology is now assessable in vivo using amyloid imaging. The first studies showed that about a third of patients with VaD due to small vessel disease or with poststroke dementia have amyloid PET imaging suggestive of AD. These new techniques will examine the interaction between vascular lesions and promotion of amyloid deposition. Although results of these on-going studies will be available in few years, these data indicate that efforts should be done in clinical practice to reduce underdiagnosis of VCI; VCI should be examined using a specific protocol which will be fully normalized soon for French-speaking patients; the sub-optimal sensitivity of screening tests prompts to use a structured interview to grade Rankin scale and to perform systematically a comprehensive assessment in stroke patients at risk of VCI; poststroke dementia occurring after 3 months poststroke may be preventable by treatment of modifiable vascular risk factors and secondary prevention of stroke recurrence according to recent recommendations.