INTRODUCTION: Thalamic nuclei are increasingly used targets for study and therapeutic modulation in drug-resistant epilepsy (DRE). Centromedian nucleus of thalamus (CMN) is relatively understudied despite known widespread connectivity and previous reports suggesting a role in both generalized and multifocal epilepsies. METHODS: The clinical and electrographic data of 13 consecutive patients who had at least one sEEG electrode sampling CMN were included. Electrographic seizures were visually analyzed for thalamic activity changes at seizure onset (SO), maintenance (SM), and end (SE) periods. Recordings were quantitatively assessed using a measure of sEEG signal interdependency (non-linear regression coefficient, h2) to compare thalamocortical synchrony at SO and SE periods. All analyses were performed using AnyWave software. RESULTS: Eighty focal-onset seizures (median 4, range 1-19 per patient) were captured and included in analysis. SOZs were mesial temporal (33%), neocortical temporal (23%), frontal (29%), and non-frontal extratemporal (15%). Median seizure length was 78 (IQR 25-133) seconds. In 74% of seizures, a change in thalamic activity was visually perceptible during the seizure with a median latency of 5 (IQR 2-9) seconds. A low voltage fast activity pattern predominantly in the alpha band was most commonly seen during SO. When thalamic activity changed during SM, a rhythmic spike-wave pattern was most often seen. Rhythmic waves predominantly in the delta band were the most common pattern of activity at SE. Average thalamocortical h2 values in the SE period were significantly higher than the SO period (Kruskal-Wallis p < 0.0001). Mesial temporal, temporal, and frontal seizures demonstrated significant increases in thalamocortical synchronization while non-frontal extratemporal seizures did not. CONCLUSIONS: Consistent dynamic ictal patterns can be visualized within CMN thalamus using sEEG recordings. Bidirectional signal synchronization peaks during the SE period and may be involved in seizure termination.