The study of sources and spatiotemporal evolution of ictal bursts is critical for the mechanistic understanding of epilepsy and for the validation of anti-epileptic drugs. Zebrafish is a powerful vertebrate model representing an excellent compromise between system complexity and experimental accessibility. We performed the quantitative evaluation of the spatial recruitment of neuronal populations during physiological and pathological activity by combining local field potential (LFP) recordings with simultaneous 2-photon Ca 2+ imaging. We developed a method to extract and quantify electrophysiological transients coupled with Ca 2+ events and we applied this tool to analyze two different epilepsy models and to assess the efficacy of the anti-epileptic drug valproate. Finally, by cross correlating the imaging data with the LFP, we demonstrated that the cerebellum is the main source of epileptiform transients. We have also shown that each transient was preceded by the activation of a sparse subset of neurons mostly located in the optic tectum. of human pathologies, including behavioral aspects of the seizure phenotype, such as locomotor patterns and loss of posture [11,13]. Up to now, locomotion represents the main readout of the epileptic phenotype and the pattern and speed of swimming behavior of larvae can indeed be measured using automated locomotion-tracking [14] to provide information on seizure severity and on the outcome of administered drugs. Electrophysiological recordings [9,[15][16][17][18] permit activity monitoring in intact larvae, but the analysis of the electrophysiology data relies mostly on their visual inspection. Moreover, it is often difficult to tell true epileptiform activity apart from physiological events, such as eye and tail movements [18] on account of the high sensibility of the electric signal to muscle activity due to the small dimension of the entire organism. Therefore, electrophysiology requires additional information for the unequivocal identification of the mutant phenotypes and for the evaluation of anti-epileptic compounds.Recently, deep learning classifiers have been employed to identify electrophysiological events [19] and, although these methods are very fast and effective, they are unavoidably affected by the need of an a priori classification of epilepsy features and by the degree of completeness of the training dataset. A high-throughput local field potential (LFP) recording platform has also recently been used to analyze high-order statistical moments for an unsupervised detection of seizures [20]. However, these electrophysiological methods for seizure identification and classification are still prone to be affected by motion artefacts. Moreover, a complete physio-pathological interpretation of LFP recordings is hindered by the impossibility of identifying the sources of the electrophysiological signals and the spatial dynamics of epileptic activity of the underlying neuronal populations.A complementary window on brain function is provided by imaging the fluorescence of ...