Abstract. Genetic divergence among conspecific subpopulations can be due to either low recurrent gene flow or recent divergence and no gene flow. Here we present a modification of an earlier method developed by Nielsen and Wakeley (2001), which accommodates a finite-site mutation model, to assess which of the two models of divergence is most likely given the observed data. We apply the method to nucleotide sequence data collected from the variable part of the mitochondrial control region in fin whales (Balaenoptera physalus) from the Atlantic coast off Spain and the Mediterranean Sea. Our estimations strongly favor a model of recurrent gene flow over a model of recent divergence and zero gene flow. We estimated the migration rate at two females per generation. While the estimated rate is high by evolutionary standards, exchange rates of this order of magnitude is low from an ecological and conservation perspective and entirely consistent with the current paucity of fin whale sightings in the Strait of Gibraltar today. Intensive commercial shore-based whaling during the 1920s removed substantial numbers of fin whales in the Strait of Gibraltar and this local population has seemingly since failed to recover. The degree of reproductive isolation among conspecifics is a basic parameter in evolutionary biology and most commonly estimated by genetic analysis. However, different evolutionary models may yield a similar degree of genetic divergence among subpopulations. For instance, low levels of genetic differentiation are commonly observed among conspecific subpopulations and could be due to either low levels of ongoing gene flow or recent divergence with zero gene flow. Methods to discriminate between these two different models have been developed by Nielsen and Wakeley (2001) based on an infinite-site mutation model. However, estimation of genetic divergence among conspecific populations is usually based on loci that evolve at relatively high rates, such as the mitochondrial control region, and where the infinitesite mutation model is inappropriate. In this study, we present a modification of the approach presented by Nielsen and Wakeley (2001), which assumes a finite-site rather than an infinite-site mutation model. We apply the approach to fin whale (Balaenoptera physalus) samples collected off the Atlantic coast of Spain and in the Mediterranean Sea.A recent study by Bérubé et al. (1998) aimed at North Atlantic and Mediterranean fin whales found a significant degree of genetic divergence (F ST estimated at 0.097) at the mitochondrial control region loci between fin whales sampled off the Atlantic coast of Spain and in the Mediterranean Sea. In contrast, significant levels of genetic divergence at six nuclear microsatellite loci were only detected among the populations furthest away from the Mediterranean Sea, such as the Gulf of Maine and western Greenland. These results suggested that an isolation-by-distance model most appropriately describes the gene flow among North Atlantic and Mediterranean Sea fin whales....