The erosion resistance of clay is an important aspect when assessing the integrity of a levee. This is because the clay layer of a dike must resist the forces that are imposed in the case of overflow or overtopping. Currently, the erosion resistance of clay in the Netherlands is being assessed based on the sand content, the liquid limit and the plasticity index. The quality of compaction is assessed by a density requirement. However, the effect of compaction on the erosion resistance of clay is not well understood and although its importance has been widely acknowledged by multiple literature sources, it has never been quantified. Therefore the goal of this thesis is to investigate what the are of compaction on the erosion resistance of clay and how large. To do this first of all a set-up has been designed and built, resembling the Erosion Function Apparatus as suggested by Briaud with some differences. Then 3 types of clay have been tested with differing initial erosion characteristics: one kind which fully complies with the current Dutch guidelines, 1 which does not quite suffice and 1 which does not suffice in any way. These 3 types of clay have been compacted at various energy levels (Modified Proctor, Proctor and 0.5*Proctor density). Also samples with differing water contents have been produced at each energy level and each clay. These samples were subsequently tested in the Erosion Function Apparatus at several flow-velocities in order to establish an erosion curve from which several erosion parameters were determined. \bigskip Having performed the erosion tests and determined the accompanying erosion parameters the results were interpreted in the Proctor curve. Plotting the results in the Proctor curve show that samples compacted at the optimum water content in the Proctor plane perform best, whereas they perform poorer the farther the water content is off from the optimum. Also it shows an increase in the compaction effort will generally result in a lower amount of erosion occurring. This result has been obtained by determining the critical shear stress, critical erosion velocity, the detachment coefficient and the velocity detachment coefficient and plotting these results in the Proctor curve. The first two parameters are measures of the amount of strength the soil has before it starts eroding at all, whereas the latter two are an increment meant to predict how much a soil might erode if the force imposed on the soil is above the erosion threshold. The results consistently showed the detachment coefficients were lowest with the samples compacted at an optimum water content and/or undergone a higher compaction effort (i.e. the soils erode less). On the other hand the critical shear stress/critical erosion velocity was highest at the optimum water content (i.e the soil erodes less), but did not necessarily increase with an increasing compaction effort. This is attributed to the loss of suction. Further, all tested soils showed their optimum erosion characteristics at or close to a degree of saturation of 85 \%. Several other methods were tried, but to no avail unfortunately. Also the rates of erosion were compared between the different types of clay. The clay which is in compliance with the Dutch codes showed the least amount of erosion, the one that did not quite suffice showed some more erosion and the clay that did not comply with the erosion guidelines at all showed the most erosion. However, it was also proved that the clay which did not quite suffice the requirements was (if compacted at optimum water content and at a sufficient energy level) only slightly more erodible than the clay which fully complied with the current Dutch regulations. Finally, the results of the tests carried out in this study were also compared to similar tests performed in the United States and the show consistent similar results. Overall, this study concludes that the use of (currently perceived) unsuitable clay as dike cover can be possible if compacted at or very close to the optimum water content and at a sufficient energy level. Also it can be concluded that compaction at the optimum water content can dramatically reduce the erodibility of a soil or even stop the erosion process. Increasing the compaction effort also reduces erodibility of a soil, although it must be kept in mind that the optimum water content also changes and thus if it is wise to do so. Finally this thesis leaves some suggestions for further research. These mainly concern the effects of cyclic wetting/drying cycles on the erodibility of the soil as well as trying to better understand the effects of the erosion process happening in the field versus the erosion process that is happening in the laboratory.