BACKGROUND: Diffusion tensor imaging (DTI) analysis along the perivascular space (ALPS) index is currently widely employed to evaluate the neurophysiological activity in various neuropsychiatric disorders. However, there remains a scarcity of studies assessing the DTI-ALPS index in individuals with methamphetamine (MA) dependence. Recent studies on animals have demonstrated a significant correlation between glymphatic activity and alterations in cerebral blood flow (CBF). Hence, investigating the DTI-ALPS index and its coupling with CBF could yield novel insights for MA-dependent patients. METHODS: In this study, we employed DTI and arterial spin labeling to investigate the ALPS index and CBF in 46 MA-dependent patients and 46 control subjects. By using DTI-ALPS, we evaluated a comprehensive diffusivity parameter that encompasses contributions from both the perivascular spaces and fiber tracts. Furthermore, a two-sample t-test was employed to assess inter-group differences. Partial correlation analysis was used to evaluate the correlations of the ALPS index with age, clinical parameters, and CBF, respectively. In addition, a causal mediation analysis was conducted to explore whether CBF mediates the causal relationship between MA-related clinical characteristics and the ALPS index. Finally, a support vector machine (SVM) was trained by the ALPS-related features and CBF features for the purpose of distinguishing MA-dependent subjects from control subjects. RESULTS: Compared to the control group, the MA-dependent group presented a decreased ALPS index, particularly in the right hemisphere. Moreover, increased diffusivities were observed along the projection fibers in the right Y-axis and the association fibers in the right Z-axis, while the AI of the diffusivity along the Z-axis association fibers decreased in patients with MA dependence. The study observed a tight coupling between the ALPS index and CBF in MA-dependent patients, and revealed significant positive correlations between the ALPS index and CBF in specific brain regions, including the right precentral sulcus, right anterior transverse collateral sulcus, left postcentral sulcus, left superior parietal lobule, left superior occipital sulcus and transverse occipital sulcus, and right temporal pole. The causal mediation analysis suggested that CBF partially mediated the alteration of the ALPS index induced by the duration of MA consumption in MA-dependent patients. Additionally, CBF/ALPS ratio was lower in the MA-dependent group compared to the controls group. An SVM trained with the ALPS-related indicators and CBF indicators achieved classification accuracy, sensitivity, specificity, and kappa values of 93.31 % ± 5.72 %, 91.56 % ± 9.14 %, 95.05 % ± 7.91 % and 86.60 % ± 11.44 %, respectively, for identifying patients with MA dependence. CONCLUSIONS: The study identified abnormal ALPS index, which has the potential to be a meaningful imaging marker for MA-dependent patients. The findings emphasized the strong coupling between the ALPS index and CBF in MA-dependent individuals, providing indirect imaging references for future research on the relationship between the glymphatic system and CBF. Moreover, the abnormal ALPS-related features and CBF features hold promise as valuable features for developing highly effective classification models.