In December 2013, a two-year-old child died from viral haemorrhagic fever in Méliandou village in the South-East of Guinea, and constituted the likely index case of a major epidemic. When the virus was formally identified as Ebola, epidemiologists started to investigate the chains of transmission, while local people were trying to make sense out of these deaths. The epidemic control measures taken by national and international health agencies were soon faced by strong reluctance and a sometimes aggressive attitude of the affected communities. Preliminary ethnographic observations were carried out by ST in Macenta (Forest region) during an assignment (October-November 2014) for the Global Outbreak and Alert Response Network (GOARN) of the World Health Organization. ST carried out participative observation, informal conversations and in-depth interviews to identify the rumours and their sources, understand the local population's perception and knowledge about the history and origin of the Ebola outbreak in Guinea. Epidemiologists involved in the outbreak response attributed the first Ebola deaths in the Forest region to the transmission of a virus by contact with fluids of patients, but other Guinean citizens believed these deaths were caused by the breach of a taboo. The epidemiological and popular explanations, mainly evolving in parallel, but sometimes overlapping, are driven by different explanatory models, a biomedical model and a traditional-religious model. The outbreak response must be flexible and must systematically document popular discourse(s), rumours, codes, practices, knowledge and opinions related to the outbreak and use this information to shape and adapt its control interventions.