Erosion control is traditionally achieved by geometrical sandtight geotextiles, with openings smaller than the grains of the subsoil to be protected. Performance tests are described in the present paper, supporting the theory that in many cases it is possible to apply geotextiles that do not meet the geometric criteria. Preliminary criteria for hydrodynamical sandtight geotextiles are derived in this paper on the bases of the test results. However some contradictory results make further research necessary.