In south-eastern Australia banding of M. schreibersii has been concentrated in four areas: north-eastern New South Wales, south-eastern New South Wales, south-eastern Victoria, and south-western Victoria and south-eastern South Australia. The present paper analyses 2083 reported movements. Only 17 of these are from one of the four areas to another with the longest movement being 810 miles. Biologically and geographically separate populations of M. schreibersii are recognized in both north-eastern and south-eastern New South Wales. Each population has its basis in dependence upon a specific nursery site which is used annually by nearly all adult females in that population. Boundaries of population ranges in New South Wales are considered to be prominent features of physiography (i.e. divides). Bats move between population ranges less often than they move within population ranges. This cannot be explained solely in terms of the distances separating roosts. Available movement records from Victoria and South Australia are consistent with the pattern described for New South Wales. Two biologically recognizable populations (i.e, different birth periods) occur in south-western Victoria and south-eastern South Australia but these may have overlapping ranges. Only one nursery colony of M. schreibersii is known from south-eastern Victoria. On present evidence it remains possible that the apparent integrity of the population associated with this nursery is merely a consequence of distance from other areas of banding activity. Detailed analyses of movements in bats may provide direct evidence as to the kinds of cues by which a given species navigates. Thus the physiographic basis described for population ranges in New South Wales is consistent with the view that M. schreibersii may orientate to waterways or divides or both. The probability that there are area differences in the subtlety or nature of navigational cues is implied by the different physiographic circumstances of south-western Victoria and south-eastern South Australia. It is suggested that knowledge of population range boundaries may aid planning of meaningful homing experiments.