There is convincing geological evidence for the historical existence of an ancient lake on the Australian–New Guinea continental shelf during the late Pleistocene. Lake Carpentaria was a vast fresh- to brackishwater lake that would presumably have provided habitat for, and facilitated gene flow among, aquatic taxa that tolerate low to moderate salinities in this region. Moreover, it has been argued that the outflow of Papua New Guinea's Fly River was diverted westward into Lake Carpentaria during this period, although this hypothesis is controversial. We predicted that these events, if a true history, would have promoted gene flow and population growth via range-expansion events in the giant freshwater prawn (Macrobrachium rosenbergii) and restricted gene flow subsequently by way of a vicariant event as sea levels rose during the late Pleistocene, and a marine environment replaced Lake Carpentaria. We tested these hypotheses using phylogeographical and phylogenetic analyses of mitochondrial DNA variation in M. rosenbergii populations sampled from the Lake Carpentaria region. Our results support the hypothesis that Lake Carpentaria facilitated gene flow among populations of M. rosenbergii that are today isolated, but contest claims of a westward diversion of the Fly River. We inferred the timing of initial expansion in the ‘Lake Carpentaria lineage’ and found the timing of this event to be broadly concordant with geological dating of the formation of Lake Carpentaria. Reconciling geological and molecular data, as presented here, provides a powerful framework for investigating the influence of historical earth history events on the distribution of biological (i.e. molecular) diversity.