The genetic variation and evolutionary mechanisms shaping Cucumber vein yellowing virus (CVYV) populations were investigated by analysis of nucleotide sequences coding for P1b, P1b/P3 and coat proteins (CP) from isolates collected in different countries. The complete genome sequence of isolate ISM from Israel was also determined and compared to those of isolates Jor from Jordan and ALM32 from Spain. This isolate had overall nucleotide identities of 94·23 and 94·96% with ALM32 and Jor, respectively. Nucleotide variation among isolates was not homogeneously distributed, with the 5′ half of the genome being more variable than the 3′ half. A Bayesian phylogenetic tree of the CP showed that CVYV isolates clustered into two main clades: isolates from the Middle East region (Lebanon, Israel and Jordan) clustered in both clades whereas the isolate from Tunisia clustered in clade I and the European isolates clustered as a homogeneous phylogroup in Clade II. A similar topology was observed for P1b but with incongruences with respect to the CP, suggesting genetic exchange among virus isolates, which were confirmed with recombination algorithms. The low genetic diversity within the European phylogroup with respect to the other isolates, neutralist tests and genetic differentiation analyses suggest that the Middle East region is the cradle of CVYV and that a unique virus introduction event occurred in Europe, where the virus spread rapidly. Taken together, these findings indicate a risk of emergence of virulent CVYV isolates in Europe either through migration from the Middle East or by genetic changes of the European isolates.