In order to better control current diseases of plants and prevent future epidemics, it is crucial to develop an improved understanding of the factors underlying pathogen emergence, adaptation and spread. Recent methodological developments in molecular epidemiology now allow tackling such question through fine reconstruction of disease dynamics in space and time. To date, essentially all studies on plant pathogens have focused on “contemporary” individuals sampled over a fairly limited period of time (around 30 years at a maximum) but recent developments in DNA sequencing technology now make possible reconstructing historical genomes dating back to previous centuries. As illustrated by the detailed reconstruction of the 19th-century potato-blight epidemic that triggered the Irish potato famine, isolation of pathogen ancient DNA (aDNA) can provide unique insight into their origins and can illuminate their phenotypic evolution, such as for instance their virulence. The aim of the MUSEOBACT project is to retrieve genomic data of historical crop pathogenic bacteria from both herbarium material and lyophilized collections. We also intend to improve the required statistical methods to reconstruct the emergence and spread of pathogens through space and time. The theoretical and empirical work will allow us to elucidate the history and evolution of some of the most devastating crop pathogenic bacteria currently in action, and to highlight the likely causes that led to their emergence and spread. As proof of concept, we attempted and succeeded in retrieving genomic data from both herbarium and bacterial lyophilized collection material. We will focus our work on several Xanthomonas pathovars and Xylella fastidiosa, two major crop bacterial pathogen complex species causing several important crop diseases and representing a huge agronomical and economical constraint in tropical, subtropical and temperate areas. The case of Xylella fastidiosa is particularly worrisome because recent emergences in Europe represent a significant threat to French and European agriculture. The major novelty of the MUSEOBACT project lies in the reconstruction of complete genomes from herbarium dried plant specimens, as this task has never been achieved for any bacterial species yet, as well as in their analyses with sophisticated and dedicated population genetics methods. This project will enable major advances in molecular epidemiology of plant infectious diseases, through both innovative molecular and statistical approaches. By i) elucidating the causes leading to crop disease emergences, ii) reconstructing pathogen invasion routes and iii) inferring key evolutionary and epidemiological parameters, our project will directly help developing new efficient disease surveillance and management strategies to prevent current/future epidemics. To ensure the success of our project, all the required skills in ancient genomics, bioinformatics, Bayesian and computational population genomics, evolutionary biology, epidemiology and phytopathology have been gathered within the MUSEOBACT scientific consortium.