To investigate genetic diversity and the population structure of the European moose (Alces alces), we analyzed 14 microsatellite loci for 694 samples collected across 16 localities. The highest genetic diversity was detected in Belarus and Russia and the lowest was found in Scandinavia. Two major genetic clusters existed, Scandinavian and continental, and some further spatial structure was detected. There was high concordance between the spatial distribution of microsatellite clusters analyzed in the present study and previously recognized mitochondrial DNA clades of moose. The split of genetic lineages calculated using approximate Bayesian computation (ABC) occurred at the beginning of the Last Glacial Maximum: approximately 29 000 and 28 000 years BP. A range-wide bottleneck detected by ABC took place 1800–1200 years BP, although a more recent decline in moose numbers was also documented in the 18th to early 20th Century. Genetic differentiation in European moose increased with geographical distance, and the Baltic Sea appeared to be a barrier to gene flow. We conclude that isolation in different glacial refugia, postglacial colonization, and declines of range and numbers in Holocene shaped the present pattern of genetic diversity of European moose. Based on genetic divergence and a lack of apparent gene flow, the contemporary Scandinavian and continental subpopulations should be treated as separate management units.