Schistosomiasis is a neglected water-born parasitic disease caused by Schistosoma affecting more than 200 million people. Introgressive hybridization is common among these parasites and raises issues concerning their zoonotic transmission. Morphological identification of Schistosoma cercariae is difficult and does not permit hybrids detection. Our objective was to assess the performance of MALDI-TOF (Matrix Assistated Laser Desorption-Ionization-Time Of Flight) mass spectrometry for the specific identification of cercariae in human and non-human Schistosoma and for the detection of hybridization between S. bovis and S. haematobium. Spectra were collected from laboratory reared molluscs infested with strains of S. haematobium, S. mansoni, S. bovis, S. rodhaini and S. bovis x S. haematobium natural (Corsican hybrid) and artificial hybrids. Cluster analysis showed a clear separation between S. haematobium, S. bovis, S. mansoni and S. rodhaini. Corsican hybrids are classified with those of the parental strain of S. haematobium whereas other hybrids formed a distinct cluster. In blind test analysis the developed MALDI-TOF spectral database permits identification of Schistosoma cercariae with high accuracy (94%) and good specificity (S. bovis: 99.59%, S. haematobium 99.56%, S. mansoni and S. rodhaini: 100%). Most misidentifications were between S. haematobium and the Corsican hybrids. The use of machine learning permits to improve the discrimination between these last two taxa, with accuracy, F1 score and Sensitivity/Specificity > 97%. In multivariate analysis the factors associated with obtaining a valid identification score (> 1.7) were absence of ethanol preservation (p < 0.001) and a number of 2-3 cercariae deposited per well (p < 0.001). Also, spectra acquired from S. mansoni cercariae are more likely to obtain a valid identification score than those acquired from S. haematobium (p<0.001). MALDI-TOF is a reliable technique for high-throughput identification of Schistosoma cercariae of medical and veterinary importance and could be useful for field survey in endemic areas.Author summary:Schistosomiases are neglected tropical diseases, affecting approximately 200 million people worldwide. They are transmitted during contact with water contaminated with the infesting stage of the parasite (the cercaria stage). Species-level recognition of cercariae present in water has important implications for field campaigns aimed at eradicating schistosomiasis. In addition, Schistosomes are able to hybridize between different species. Identification of Schistosomes cercariae on microscopy is difficult because of their similarity, and it does not allow hybrids to be distinguished. Molecular biology techniques allow a reliable diagnosis but are expensive. MALDI-TOF is a recent technique that permits an inexpensive identification of micro-organisms in a few minutes. In this paper, we evaluate MALDI-TOF identification of Schistosomes cercariae. We have implemented a database of MALDI-TOF cercariae spectra obtained from parental strains and hybrids of species of medical or veterinary interest, allowing reliable identification with an accuracy of 94%. The identification errors mainly come from confusion between the natural Corsican hybrid (S. haematobium x S. bovis) and S. haematobium. The use of machine learning algorithms permits to obtain an accuracy of more than 97% in the recognition of these two parasites. In conclusion, MALDI-TOF is a promising tool for the identification of Schistosome cercariae.