Epilepsy prevention is an urgent and global unmet need today. Particularly, in structural epilepsies that may occur months or even years after an epileptogenic insult (e.g., stroke, tumor, trauma, or status epilepticus), this latency time provides a window for preventive treatments for individuals at risk of developing epilepsy. In this regard, the identification of biomarkers to detect the subgroup of individuals with an increased risk of epilepsy after the epileptogenic insult is essential for the development of preventive treatments.A multicenter study was conducted using four different models of epileptogenesis in rats; plasma sampling at day 2 and day 9 during the latency period from animals that did or did not develop epilepsy during long-term video-electroencephalogram monitoring was performed to identify the profiles of differentially expressed circulating miRNAs and isomiRs as biomarkers of epileptogenesis. The miRNAs and isomiRs were identified and measured using a genome-wide small RNA sequencing platform and were subjected to ROC analysis to determine the performance of putative biomarkers. At 2 days after the insult, several miRNAs and isomiRs were identified in plasma that predict epileptogenesis in a model-specific manner. One miRNA, miR-3085, showed good sensitivity as a prognostic biomarker of epileptogenesis in all four models (AUC 0.729, sensitivity 83%, specificity 64%, p<0.05).The identified miRNAs and plasma isomiRs are mostly etiology-specific biomarkers rather than common prognostic biomarkers of epileptogenesis. Importantly, circulating miRNAs, such as miR-3085, with a high negative predictive value of epileptogenesis in different etiologies, could be useful for screening individuals at risk for epilepsy.