Objectives: To evaluate the utility of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in predicting the response of chemotherapy and clinical outcomes in primary central-nervous-system lymphoma (PCNSL) patients. Methods: DCE-MRI in 56 patients enrolled in a prospective study was performed at baseline and 30 days after treatment from 2016 to 2019. Multivariate logistic regression analyses were performed to assess risk factors for tumor responses. The predictive values of related parameters derived from DCE were analyzed via receiver operating characteristic (ROC) curve analysis. To evaluate prognostic factors, the Kaplan–Meier survival analysis with log-rank tests and Cox regression tests were analyzed. Results: Ktrans and Ve were higher in the non-response group than in the response group (p < 0.05). The Ktrans and the percentage of Ktrans decreased after 30 days of treatment were independent predictors of chemotherapy responses (p = 0.034 and p = 0.019). ROC analysis indicated that the cut-off point of Ktrans for predicting chemotherapeutic responses was 0.353 min−1 (AUC, 0.941; 95% CI, 0.87–1; p < 0.001) and percentage of Ktrans decreased after 30 days of treatment was 15.2% (AUC, 0.858; 95% CI, 0.742–0.970; p < 0.001). The greater decrease in Ktrans correlated with a longer progression-free survival (PFS) (χ2 = 13.203, p < 0.001). The higher Ktrans was an independent predictor for shorter PFS (hazard ratio, 10.182; 95% CI, 2.510–41.300; p = 0.001). Conclusions: Ktrans and Ktrans change measured by DCE-MRI were reliable biomarkers for predicting chemotherapy responses in PCNSL patients. Key Points: • Baseline Ktransand greater decrease in Ktranscan predict chemotherapeutic efficacy. • DCE-MRI provides quantitative parameters reflecting the tumor microenvironment. • Targeted treatment therapy can be given with more evidence in the future. © 2020, European Society of Radiology.