Numerous epilepsy-related genes have been identified in recent decades by unbiased genome-wide screens. However, the available druggable targets for temporal lobe epilepsy (TLE) remain limited. Furthermore, a substantial pool of candidate genes potentially applicable to TLE therapy awaits further validation. In this study, we reveal the significant role of KCNQ2 and KCNQ3, two M-type potassium channel genes, in the onset of seizures in TLE. Our investigation began with a quantitative analysis of two publicly available TLE patient databases to establish a correlation between seizure onset and the downregulated expression of KCNQ2/3. We then replicated these pathological changes in a pilocarpine seizure mouse model and observed a decrease in spike frequency adaptation due to the affected M-currents in dentate gyrus granule neurons. In addition, we performed a small-scale simulation of the dentate gyrus network and confirmed that the impaired spike frequency adaptation of granule cells facilitated epileptiform activity throughout the network. This, in turn, resulted in prolonged seizure duration and reduced interictal intervals. Our findings shed light on an underlying mechanism contributing to ictogenesis in the TLE hippocampus and suggest a promising target for the development of antiepileptic drugs.