From a prospective registry of all consecutive patients with a supratentorial ischaemic stroke, those with a compatible CT lesion were selected to study topographical relationship, clinical syndrome, vascular risk factors, signs of large-vessel disease or cardiogenic embolism, and mortality in cases with an infarct in the anterior choroidal artery (AChA) territory in comparison with other infarct subtypes. First we identified the area supplied by the AChA: in accordance with the consensus in the literature the posterior two-thirds of the posterior leg of the internal capsule was considered as certain AChA territory. After reviewing CT scans, all presumed small deep AChA territory infarcts were displayed in a schematic composite picture of super-imposed areas of infarction in different shades of grey. Infarcts that were located largely outside the generally included territory were presumed to belong to a different vascular territory. Thus, 77 small deep infarcts were considered to be located within, and 83 outside the AChA territory. Twenty-nine AChA infarcts extended from the internal capsule upwards into the posterior paraventricular corona radiata region. Furthermore, the composite representation of 26 infarcts restricted to the posterior part of the paraventricular corona radiata region showed almost complete overlap with the area occupied by AChA infarcts that extended upwards. We therefore concluded that the posterior paraventricular area is most likely supplied by the AChA. The frequency of a clinical lacunar or a cortical syndrome did not differ between small deep AChA and remaining small deep infarcts. Comparison of vascular risk factors by way of multivariate regression analysis only showed that a significant carotid stenosis was more frequent (adjusted odds ratio 8.87; 95% confidence interval 1.44–54.50), and a cardioembolic source was less frequent (odds ratio 0.24; 95% confidence interval 0.07–0.92) in AChA infarcts than in the other small deep infarcts. Carotid stenosis and cardiac embolism were less frequent in AChA infarcts than in superficial infarcts (odds ratio 0.33, 0.23, respectively; 95% confidence interval 0.15–0.74, 0.09–0.52, respectively). One month and one year mortality were lower in small deep infarcts compared with superficial infarcts, but most favourable in the AChA group. However, this was probably related to younger age in the AChA patients. Larger AChA infarcts were infrequent in our series; six of such cases did not differ in any respect from superficial infarcts. We conclude that the posterior paraventricular corona radiata region is most likely supplied by the AChA, and that AChA infarcts do not constitute a separate brain infarct entity. Consequently, such infarct cases should be treated as similar brain infarct subtypes.