Abstract A comparison of Blaugrund and Monte Carlo centroid-shift calculations is presented. We find that the two methods of calculation agree to within 15% if the lifetime is greater than several times the flight time between collisions. The differences are systematic and hence predictable. These discrepancies depend only weakly on the reduced energy parameter ϵ and are a function of the mass ratio only when the projectile-target mass ratio is close to unity. Furthermore, the discrepancies between these methods of calculation only show a weak dependence on the choice of atomic interaction potential. On the other hand, for lifetimes that are shorter than a few times the flight time between collisions, the situation is more complex.