Objective: To compare MRI using perfusion and diffusion techniques with 6-[ LGGs) and to identify the best imaging parameter to differentiate patients with different prognosis. Methods: Between 2010 and 2015, 12 patients with a pathology-proven diagnosis of LGG and MR (with perfusion and diffusion sequences) and a PET study during their follow-up were retrospectively included in our study. Cerebral blood volume (CBV) and apparent diffusion coefficient (ADC) maps on MR studies and PET images were evaluated using a region of interest-based method. All patients were categorized as stable or as having progressive disease at 1-year follow-up. Statistical analysis was performed using Pearson's correlation test and multivariate analysis of variance (p , 0.05). Results: No significant correlations were found between PET parameters [maximum tumour-to-controlateral normal brain ratio (T/N max ) and tumour-to-striatum ratio] and ADC or relative CBV values measured in both PET hotspot regions and areas of maximum signal alterations. T/N max demonstrated a good sensitivity (83%) and specificity (100%) for differentiating two subgroups of patients with different outcomes at 1-year-follow-up (p , 0.05). Conclusion: Perfusion and diffusion MR images provide different information compared with 18