Background Temporal lobe epilepsy (TLE) is characterized by alterations of brain dynamic at large scale associated with altered cognitive functioning. Interindividual variability of brain activity is a source of heterogeneity in this disorder. Here, we aimed at analyzing dynamical reconfiguration of brain activity, using the neural fingerprint approach, to delineate subject-specific characteristics and their cognitive correlates in TLE. Methods We collected 10 minutes of resting state electroencephalography (128 channels), free from epileptiform activity, from 68 TLE patients and 34 healthy controls. The functional network was defined by the spatio-temporal spreading, across cortical regions, of aperiodic bursts of activations (neuronal avalanches). This metric allowed encapsulating brain reconfiguration patterns into the avalanche transition matrix (ATM). We used a neural fingerprint approach to differentiate across controls and patients diagnosis, linking altered brain dynamic with cognitive outcome. Findings Patients’ brain dynamics were more stereotyped as compared to controls. The neural fingerprint using ATMs differentiated, in a data-driven fashion, patients with respect to healthy controls, being sensitive not only to the pathology but also to the subtype (unilateral vs. bilateral TLE). Notably, in unilateral TLE patients, a better memory performance was associated with a larger similarity of the brain dynamic configuration with controls. Interpretation TLE is characterized by reduced variability and more stereotyped brain dynamics, implicating widespread alterations across the brain. These alterations correlate with cognitive function in patients with unilateral TLE. This study underscores the utility of brain fingerprinting in elucidating disease-specific brain dynamics, offering novel metrics for personalized patient care.