Recent investigations have shown that an increase of the packing density of armour blocks on a slope usually leads to an increased stability. This is also the case for cubes. Test on slopes with cube armour units with different values for the porosity yield an increase of stability until a certain optimum packing density. As a first conclusion this would imply that breakwaters with cubes should be designed using this optimum packing density. However, the various packing densities of the cubes in the laboratory were realised by placing the cubes carefully by hand in the model breakwater. In reality these cubes will be placed by dropping them from and above the water surface. The exact position for dropping the blocks (either from a crane or from a side dumping vessel) can be determined easily, also in a real work under execution using standard GPS equipment. However, during the dropping process the blocks will not arrive exactly below the point of release. When touching the slope the cubes will tend to slide down towards the already placed blocks. Consequently at that moment the blocks will tend to be placed as a revetment. On the other hand, in case the block hits the blocks already placed, the new block will either turn the other way or it will jump away on the impact. In the most "unfortunate" case, the block will stand on one of the corner points which differs 45 from the slope of the secondary armour. This will result in a considerably higher porosity of the armour layer. The process of falling is a stochastic one. The average "landing-location" is exactly below the point of release, but there is a considerable standard deviation. In the past research has been done to find this standard deviation for rock, blocks and other elements to be dropped on a horizontal bed. However this was done for relatively deep water (where block reached the stationary falling velocity). In order to investigate this effect, in a basin model cubes (DN =0.15 m) have been dropped according to various patterns on a 2:3 slope. The result is a rather "random" placement of the cubes on the slope. The scale was chosen to avoid scale effects due to impacts. The results depend very much on the initial position of the blocks at the toe of the breakwater. Also we concluded that the packing density depends much on the distance between the individual release points. In case the drop-distance is too small (with as purpose to achieve a very high packing density), the result was in many cases a considerably lower packing density because of an increased layer thickness. Basically two test series have been done, one series had the main axis of the blocks parallel to the crest of the breakwater, while the other series had a shift of 45 . In both cases the distance between the blocks is quite sensitive. The results of all test are being analysed and result in an "optimum" drop distance both in x as well as in y direction (x is parallel to the crest line, y is perpendicular to the crest line, both measured in a horizontal plane). A small drop distance in some cases will result in a dense packing, but there is a considerable risk that the packing becomes much less dense then anticipated.