Objective:To investigate the value of diffusion tensor imaging (DTI) in evaluating the cognitive function of patients with leukoaraiosis (LA).Methods:A prospective study was chosen. Sixty patients with LA admitted to our hospital from July 2019 to May 2021 were selected. All patients accepted brain MRI. According to Fazekas visual grading standards, these 60 LA patients were divided into mild LA group, moderate LA group, and severe LA group ( n=20). Mini-mental state examination (MMSE) was used to evaluate the cognitive function of these patients. Diffusion tensor imaging was used to analyze the fractional anisotropy (FA) values of 6 regions of interest, namely the frontal lobe, temporal lobe, occipital lobe, parietal lobe, anterior horn of lateral ventricle, and corpus callosum. The differences of general data, cognitive function and FA values among the 3 groups were compared. The cognitive functions were assessed based on FA values (FA scores), and compared with those evaluated by MMSE (clinical scores). Results:(1) Age and homocysteine (Hcy) level of patients in mild, moderate and severe LA groups were increased successively, with statistical differences ( P<0.05). As compared with that in the mild and moderate LA groups, the educational level of patients in the severe LA group was significantly lower ( P<0.05). (2) Memory scores of patients in mild LA group, moderate LA group and severe LA group decreased successively, with statistical differences ( P<0.05). As compared with mild LA group and moderate LA group, MMSE total scores and recall scores in severe LA group were statistically decreased ( P<0.05). (3) FA values of parietal lobe and corpus callosum were successively decreased in mild, moderate and severe LA groups, with significant differences ( P<0.05). As compared with those in the mild and moderate LA groups, the FA values of frontal lobe, temporal lobe, occipital lobe and anterior horn of lateral ventricle in severe LA group were significant decreased, ( P<0.05). (4) There were no significant differences in FA scores and clinical scores of cognitive function scores in mild LA group ( P>0.05), except for attention and counting scores. There were no significant differences in FA scores and clinical scores of cognitive function scores in moderate LA group ( P>0.05), except for language function scores and tMMSE otal scores. Except for scores of location orientation, attention and counting, FA scores and clinical scores of all cognitive function scores in the severe LA group were significantly different ( P<0.05). Conclusions:The LA severity is related to the patient's age, Hcy level, education level, and cognitive function. In patients with mild LA, the cognitive function can be assessed according to FA values in addition to attention and counting ability; in patients with moderate LA, cognitive function can be assessed according to FA values in addition to language function; the cognitive function of patients with severe LA cannot be assessed according to FA values.