Objective To analyze the predictive value of diffusion tensor imaging (DTI) for the early diagnosis and clinical outcome in patients with acute cerebral infarction (ACI). Methods The clinical data and follow-up data of 80 patients with unilateral ACI in the hospital from September 2017 to June 2018 were collected retrospectively. There were hyperacute phase (onset time less than 6h) of 16 cases, the acute phase (6 to 72h) of 38 cases and the subacute phase (72h to 14d) of 26 cases, all of them were given MRI examination, including regular sequence, DTI sequence with at least 3 months of followup, clinical outcomes were assessed according to the National Institutes of Health Stroke Scale (NIHSS) score. Results A total of 80 patients with ACI could find the infarct lesions by conventional MRI and DTI with 100% detection rate. Infarcts lesions showed high signal on T2WI and DWI images, and mainly showed low signal on DTI. The FA and DCavg values of the lesions were lower than those of the healthy side (P<0.05). The FA and DCavg values at the lesion side of patients in different onset periods were lower than those at the healthy side. Arranging the FA and DCavg values from large to small, the order was hyperacute phase, acute phase and subacute phase (P<0.05). After 3 months of follow-up, it was found that arranging the reduction rate of FA and DCavg values in the lesion area from small to large, the order was super acute phase < acute phase < subacute phase, and NIHSS score was hyperacute phase, acute phase and subacute phase (P< 0.05). Pearson correlation analysis found that the decrease rate of FA and DCavg values in the lesion area of ACI patients was positively correlated with NIHSS score (r=0.465, 0.524, P<0.05). Conclusion DTI parameters have specific changes direction in ACI patients at different onset period, which can be used as a basis for clinical diagnosis. FA and DCavg values are closely related to prognostic indicators and has the potential to be used to predict the clinical outcomes.