Background: In 1989, Louis Caplan first used the term branch atheromatous disease (BAD) to describe an occlusion or stenosis at the origin of a deep penetrating artery of the brain, associated with a microatheroma or a junctional plaque, and leading to an internal capsule or pontine small infarct. BAD remained an understudied concept for decades. In recent years, the increasing diffusion of high-resolution magnetic resonance imaging (HRMRI) techniques brought new attention to the BAD debate. We have reviewed clinical studies dealing with BAD-related stroke checking whether a univocal definition of BAD existed, as well as to what extent were consistently associated clinical and imaging features reported. Summary: We conducted a search of the available literature published up to October 20, 2015 via PubMed using the following search terms: ‘branch atheromatous disease,' ‘intracranial branch atheromatous disease,' ‘cerebral branch atheromatous disease,' combined with ‘stroke.' Forty-six articles were included. We found discrepant definitions and a large variation among clinical features reported in BAD-related stroke patients: among others, a consistent association between BAD and any specific vascular risk factor profile was not detected. Despite this, early neurological deterioration (END) was consistently reported to occur frequently in such patients, although no clear-cut rate range or specific predictor or mechanism of progression was established. In a majority of the studies reporting imaging data, BAD diagnosis was not based on the selective site or type of arterial walls changes, but was inferred based on the vascular territory, size and/or shape of the ischemic lesion. Following the concept that these changes are seated proximally along the perforator artery, differently from to lipohyalinosis changes located distally, the consequent ischemic lesion was hypothesized to be larger in BAD than in lacunar infarcts. However, across reviewed studies, there was little consistency on the dimensional cutoff used to define BAD-related infarcts. In the last few years, a still limited number of studies using HRMRI techniques is providing preliminary proofs that atheromatous changes causing selective remodeling in the parent vessel and extending through the proximal segment of perforating vessel may subtend BAD. Key Messages: Our literature search showed the lack of a clear-cut definition of BAD, although BAD-related strokes were consistently considered a high risk of END. The use of high-resolution imaging techniques in the assessment of small subcortical strokes may represent the cornerstone in the perspective to better delimiting the boundaries of BAD as a nosological entity.