While empirical research on attitudes towards languages and linguistic varieties has become increasingly popular from the 1960s onwards (e.g. Lambert, Hodgson, Gardner, & Fillenbaum, 1960), experimental investigations into the ability to correctly identify the origin of speakers are in comparison still relatively scarce. One knows that the ability to correlate a stretch of uncategorised speech (token) with a series of models (types) is experientially acquired in early childhood (e.g. Kristiansen, 2010), but how similar are those abilities in adulthood and across European nations? English as a Lingua Franca (ELF) has become an integral part of the linguistic reality in Europe (and of the linguistic scenario in the entire world) (e.g. Jenkins, Baker, & Dewey, 2018). Whenever people communicate with anyone who is not a speaker of their own native language in any European country, most of the time they communicate in English. But does their L1 accent still shine through? Will they be recognised (and in most cases probably also stereotypically judged) on the basis of just a short stretch of speech when they communicate in ELF? Part I of this paper outlines the design of the first large-scale pan-European project on L1 and L2 identifications of ELF in Europe, including 785 respondents from 8 countries. Exploratory analyses confirm the hypothesis that statistically significant asymmetries would show up across different European countries or regions. Part II of this paper aims to explain these asymmetries through a multifactorial statistical analysis (Geeraerts, Grondelaers, & Speelman, 1999; Tagliamonte & Baayen, 2012; Speelman, Heylen, & Geeraerts, 2018).