One of the fundamental goals of phylogeographical studies should be to achieve a comprehensive geographical sampling of any investigated group. In this study, we conducted the most comprehensive geographical investigation to date for the great spotted woodpecker complex (Dendrocopos major), including populations from North Africa and Eurasia [including specimens from China, Japan and southern Caucasia (Anatolia, Azerbaijan and Iran)], in order to evaluate its genetic structure and population history. At the same time, we tested species limits within the D.?major complex, which currently includes 14 recognized subspecies based on morphology and coloration. We based our study on haplotypes for the mitochondrial gene NADH dehydrogenase subunit 2 (ND2). Most haplotypes were obtained from museum toe pads, although we also used some previously published data. We also tested gene flow through MDIV, and estimated divergence dates among lineages using BEAST. The analysis of 352 base pairs of the ND2 gene from 155 individuals sampled from 33 populations showed significant phylogeographical structure across the breeding range. Our results found four distinct and reciprocally monophyletic clades: China, Japan, IranAzerbaijan and EurasiaNorth Africa, with no phylogeographical structure within them. Coalescent-based gene flow analysis showed restricted gene flow between China and Japan and between Japan and Eurasia. On the basis of the gene flow and phylogenetic analysis results, we propose the recognition of at least four different species within the complex. We also propose that, within the EurasiaNorth Africa clade, a rapid population expansion through leading edge expansion from refugia in Iberia, Kursk and North Africa, as well as irruptive and loop migration, can explain the lack of phylogeographical structure. (c) 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, , .