The effect of genetic variability as indicated by allozyme heterozygosity on developmental homeostasis as expressed by fluctuating morphological asymmetry (FA) is under current debate. Recent studies of brown hares Lepus europaeus Pallas, 1778 in Austria (Hartl et al. 1995) revealed a negative relationship between FA and allozyme heterozygosity in non-metric but not in metric skull characters. We aimed to find out whether non-metric traits of brown hares are generally more prone to FA than metric, or whether metric traits also show increased FA in populations with drastically reduced genetic variability. We studied variation in 34 enzyme systems and the relationship between overall individual heterozygosity based on polymorphic loci and FA in 27 non-metric and six metric bilateral skull characters of 96 brown hares from three populations in Britain and two in New Zealand. All brown hares in New Zealand are considered to be descendents of only six founding individuals from Britain and were expected to have considerably reduced gene pool variability. Only six polymorphic loci were found. Allozyme heterozygosity in the New Zealand populations was not significantly lower than in British populations. However, both the New Zealand and the British populations had significantly lower genetic variability than the Austrian populations studied earlier. This suggested a (historic) genetic bottleneck in British brown hares having preceeded the one in hares in New Zealand. It corresponds to the hypothesis of the deliberate introduction of brown hares to Britain by the ancient Romans. Neither at the individual nor at the population level was there any significant relationship between FA and allozyme heterozygosity. Despite the significantly reduced genetic variability in hares from New Zealand and Britain, their population-specific levels of non-metric and metric FA were significantly lower than in hares from Austria. This might result eg, from genetic drift following the founding event in Britain, or from stronger selection against hares with low developmental homeostasis in Britain and New Zealand than in Austria, due to environmental differences. The study demonstrates that in brown hares reduced genetic variability is not necessarily connected with low levels of developmental homeostasis as the earlier study of Austrian populations might have suggested. Rather, the particular phylogeny of populations has to be taken into account, when interpreting relationships between developmental homeostasis and genetic variability.