In a recent Commentary in this journal, Pamilo (2004) criticized our analysis of the spatial genetic structure of the Eurasian lynx in Scandinavia (Rueness et al. 2003). The analyses uncovered a marked geographical differentiation along the Scandinavian peninsula with an apparent linear gradient in the north–south direction. We used computer simulations to check on the proposition that the observed geographical structure could have arisen by genetic drift and isolation by distance in the approximate 25 generations that have passed since the last bottleneck. Pamilo disapproved of our choice of population model and also how we compared the outcome of the simulations with data. As these issues should be of interest to a wider audience we discuss them in some detail.