Particle trapping in nine different grassed swales was measured successfully with a standardised runoff event simulation procedure. The percentage of total suspended solids removed ranged from 79 to 98%. It was found that sedimentation processes, rather than grass filtration governed the overall particle trapping efficiency. The highest particle trapping efficiency was observed in the field swales with dense, fully developed turf. A high infiltration rate was beneficial for the particle trapping and an increased swale length made it possible for smaller particles to be captured. A densely vegetated, ten metre long swale, receiving a stormwater flow of 1.0 litres per second, may capture a majority of the waterborne particles with settling velocities larger than 0.1 metres per hour. A simple model of particle trapping efficiency in grassed swales was developed and tested. It was found that mean swale residence time could be used as a design parameter for particle removal in grassed swales. The suggested exponential relationship between mean swale residence time and particle settling velocity associated with a certain trapping efficiency is so far only valid for a limited range of swale designs and residence times.