The particulate matter emitted by transports, in particular Diesel vehicles, has been considered toxic for many years, and is subject to world-wide emission standards. Recently, size distribution and particulate number has emerged as influential parameters of particulate toxicity.Over the past twenty years, many technologies have been studied for the removal of Diesel particulates. These technologies include numerous kind of traps, that are efficient mainly for solid particles. The work presented here has been devoted to the study of the influence of various traps on particulate size distribution and number, by means of SMPS (Scanning Mobility Particle Sizer) technique, in the exhaust of a VW 1.9 I TDI engine set in an engine test cell. The candidates tested are the well known Corning wall-flow monolith, the 3M wound ceramic fibre cartridge, the Buck knitted fibre cartridge and the HJS/SHW sintered metal trap. Prior to the trap comparison, some preliminary tests highlight the need for additional dilution air in the sampling line in order to avoid agglomeration and condensation upstream of the measurement device. Then, engine running conditions are shown to have a low influence on particulate size distribution. Traps appear to affect the size distribution in quite different ways during the filtration phase or the regeneration. In particular, wall flow traps are shown to create fine mode particles in some engine and particulate loading conditions, whereas deep bed filters present varying filtration efficiencies during the loading process. During regeneration, particulate re-emissions appear in all cases.