The behaviour of clayey soils is strongly correlated to their microstructure and evolution thereof. Microstructural investigations have contributed to understanding soil behaviour and have supported the development of multi-scale coupled models. One of the most accessible methods to characterise soil microstructure is mercury intrusion porosimetry, which provides a pore-size distribution (PSD) ranging from few nanometers to several hundreds of micrometers. PSDs can be used to compute micro and macro strains or simply to estimate, in aggregate microstructures, the void ratios associated with macropores and micropores. However, in both cases, a boundary has to be set to separate the different pore populations. This paper discusses some criteria that have been proposed in the literature to separate the pore populations. The discussion is illustrated with extensive micro structural data obtained for Maryland clay. The paper highlights the effect of initial conditions and boundary conditions on the delimiting diameter given by some of the criteria. It is also shown that using different criteria will yield different values of delimiting pore size, with a risk of obtaining unrealistic estimates of micro and macro void ratios and strains. Finally, it is suggested to account for known soil behaviour to interpret microstructural data.