Abstract Two methods were developed to investigate the porosity of moving slugs in situ during horizontal slug flow pneumatic conveying. The first method consists in applying a permeability model in combination with measurements of pressure loss and fluid velocity along the slugs. A review of existing models describing the resistance of porous structures to fluid flow revealed that the semi-empirical model of Ergun is particularly suitable to investigate the porosity profile along moving slugs. The second method consists in a direct determination method involving a slug-catcher able to catch a moving slug in a fraction of a second and simultaneously separate it into three horizontal layers. Those two methods were applied to analyse the porosity of naturally occurring slugs during pneumatic transport of polypropylene pellets. It was found that in contrast to common belief, slugs are slightly fluidised structures that do not display any porosity gradient over the pipe cross-section height. The slug porosity appeared independent of the gas conveying velocity, all slugs displaying an average porosity around 0.41, which is slightly higher than the bulk porosity of 0.38. Most of the slugs displayed a rear that is denser than the front. However, some slugs had a front that is denser than the rear while other slugs displayed a relatively constant porosity over the entire length. Those unique results refuting the commonly used hypothesis that slugs are compact structures give a new incentive to the area of slug flow pneumatic conveying. While bulk solids mechanics can no longer be applied to explain the stresses induced by moving slugs, the validity of other theories that imply that slugs are fluidised structures should be investigated.