Highly pathogenic avian influenza (HPAI) H5N1 viruses were introduced to Hungary during 2006-2007 in three separate waves. This study aimed at determining the full-length genomic coding regions of the index strains from these epizootics in order to: (i) understand the phylogenetic relationship to other European H5N1 isolates, (ii) elucidate the possible connection between the different outbreaks and (iii) determine the putative origin and way of introduction of the different virus variants. Molecular analysis of the HA gene of Hungarian HPAI isolates obtained from wild birds during the first introduction revealed two groups designated Hungarian1 (HUN1) and Hungarian2 (HUN2) within sublineage 2.2B and clade 2.2.1, respectively. Sequencing the whole coding region of the two index viruses A/mute swan/Hungary/3472/2006 and A/mute swan/4571/Hungary/2006 suggests the role of wild birds in the introduction of HUN1 and HUN2 viruses: the most similar isolates to HUN1 and HUN2 group were found in wild avian species in Croatia and Slovakia, respectively. The second introduction of HPAI H5N1 led to the largest epizootic in domestic waterfowl in Europe. The index strain of the epizootic A/goose/Hungary/14756/2006 clustered to sublineage 2.2.A1 forming the Hungarian3 (HUN3) group. A common ancestry of HUN3 isolates with Bavarian strains is suggested as the most likely scenario of origin. Hungarian4 (HUN4) viruses isolated from the third introduction clustered with isolate A/turkey/United Kingdom/750/2007 forming a sublineage 2.2.A2. The origin and way of introduction of HUN4 viruses is still obscure, thus further genetic, phylogenetic, ecological and epidemiological data are required in order to elucidate it.