A Bayesian unfolding has been applied to ionisation distributions due to 5.4 MeV alpha particles in a 20-nm site obtained using Monte Carlo simulations, taking into account different detection efficiency conditions. The ideal case of a target volume with uniform efficiency has been investigated to study the dependence of the reconstruction on prior distributions. A simplified approach has been used to treat the case of a target volume with non-uniformly distributed efficiency, like the sensitive volume of the track-nanodosimetric counter. Our results point out that Bayesian analysis provides a effective tool for reconstructing the true ionisation distributions, well beyond the maximum measured cluster size.