A total of 469 brown hares Lepus europaeus Pallas, 1778 from 20 sampling sites in Austria were examined for genetic diversity within and among populations by means of horizontal starch gel electrophoresis. Fourteen out of 54 presumptive structural loci were polymorphic, one of which was excluded from further population genetic analyses due to the occurrence of a null-allele. The mean proportion of polymorphic loci (P) was 15.3% (SD 2.2%), and mean expected average heterozygosity (H-e) was 4.6% (SD 0.5%). Both relative (F-ST = 5.4%) and absolute (mean Nei's 1978 D = 0.0016, SD 0.0016) genetic differentiation among populations were low, suggesting a generally high level of migration. Cluster analysis revealed some separation of brown hare populations in western and northern Austria from those in the east and in the south. In 131 individuals, mtDNA was digested with a battery of 16 restriction endonucleases. Besides the standard type I which occurred exclusively in most of the populations, five additional haplotypes, each of them deviating from type I by one base pair substitution, were detected. Together with rare alleles at allozyme loci, the distribution of variant haplotypes corroborated the spatial pattern obtained by allozyme distances and suggested considerable immigration of brown hares from the adjacent countries in the east and south. Twenty non-metric skull traits were scored in 443 individuals. Character variants were dichotomized (0/1) and the respective frequencies were used to calculate C. A. B. Smith's 'mean measure of divergence' (MMD) among five population groups. Morphological differentiation was in accordance with the major population genetic pattern as revealed by molecular techniques. MtDNA variation (nucleon diversity, nucleotide diversity) and morphological variation (mean of SD in single characters) within populations were not significantly associated with one another, and did not show a relationship with indices of genetic variation obtained by allozyme analysis. These findings suggest that variability in only one of these characters cannot be considered representative for overall gene pool diversity-within populations.