Zieria prostrata (Rutaceae) is known from only four headlands within a 3-km stretch of coastline in New South Wales, Australia. The species was presumed to have occurred at a headland 24 km south of its present range. We used random amplified polymorphic DNA analysis to assess patterns of genetic variation within and among the extant populations. The analysis also included an individual reputedly rescued from the now extinct population. A high level of population divergence was revealed by principal coordinate analysis and an analysis of molecular variance (AMOVA; 37% among populations). Our genetic findings provide implications for the conservation management of the species. First, the loss of any one population would lead to a severe loss of genetic variation. Second, an adequate ex situ collection must sample the full range of genetic diversity from all populations. Third, the consequences of mixing populations may be an important conservation consideration if further translocations proceed. Fourth, the individual apparently sampled prior to its population extinction is genetically similar to individuals from one of the extent sites. This degree of similarity was unexpected and, after further investigation, led to the conclusion that prior existence of the species at the site is doubtful. Subsequently, a planned reintroduction program was abandoned. So far, of these four management implications, only the last has had a direct management outcome. These implications that failed to lead to practical management outcomes did so because the same management recommendations could be obtained without genetic research. Clearly, the challenge for more effective conservation managers. This may be best achieved by assessing the outcomes of genetic studies already conducted.