Although the global burden of measles has been substantially reduced since the introduction of the first measles vaccine in the 1960s, large outbreaks continue to affect populations in every WHO region. Even in countries with a high national vaccine uptake, social and spatial heterogeneity in coverage lead to under-immunised populations, where the importation of cases can cause large transmission clusters. This thesis explores how local transmission risk can be identified using different data sources: i) routinely collected individual-level case surveillance data, and ii) population-level factors such as vaccination coverage and recent outbreaks. In the absence of regular sub-national serological surveys, transmission trees from previous outbreaks can be used to identify areas repeatedly associated with transmission events. I developed the R package o2geosocial to reconstruct who infected whom from routinely collected surveillance data, and to compute the number of cases per transmission cluster, i.e. the cluster size distribution. This method infers the infector, infection date and importation status of each case using their onset date, location, age, and genotype. In the first chapter of the thesis, I outlined the methodology implemented in the package and applied it to simulated local outbreaks. The method was able to reconstruct the simulated transmission dynamics and highlighted regions repeatedly associated with secondary transmission. In the second chapter, I applied o2geosocial to data from the national measles database in the United States, which lists cases reported between 2001 and 2016. Both studies illustrated the ability of this method to reconstruct transmission history from widely collected epidemiological information in a variety of contexts and geographical scales. Countries become eligible for certification of World Health Organisation’s measles elimination status after national transmission is interrupted for three years in the presence of high national vaccine coverage. Recent major outbreaks in countries where measles had been declared eliminated (e.g., United Kingdom, Brazil, Greece) illustrate that current indicators of elimination may be imperfect predictors of outbreak risk. In the third and fourth chapters of this thesis, I studied the impact of recent levels of local incidence and vaccine uptake on the risks of importation, cross-regional and local transmissions by implementing a time-series model using the R package surveillance. I applied this model using the daily number of cases reported in France between 2009 and 2017 and discussed how local indicators can inform the risks of national outbreaks.