Subsea rock installation is often used in offshore engineering. With offshore activities in increasing water depths a flexible fallpipe is currently used to accurately install the rock particles. The total required volume of rock material during installation can be determined by summing the geometrical volume, the operational losses and the displacements of the seabed. Accurate knowledge of the volume loss during installation is needed to ensure proper project management and cost estimation. This thesis provides insight in the processes which influence the volume loss during subsea rock installation and focusses specifically on immediate particle penetration. During installation the following processes affect volume loss: - Immediate deformation of the seabed - Loss of fine particles due to deep sea currents - Immediate particle penetration in the seabed - Flow of cohesive material in to the pores of the rock fill - Long term settling of the berm due to consolidation - Possible erosion of the seabed - Stones deposited outside the theoretical perimeter of the construction With a lot of oil and gas activity offshore Norway the problem of large volume losses becomes more important since this seabed consists of very soft clay. Additionally, since the rocks are installed in deeper waters using a flexible fallpipe, the impact velocity at the bottom is high (~3 m/s). Due to the high velocity and the low strength of the soil, the particles are thought to penetrate relatively deep into the soil (~0.1 m). In this thesis a model is presented to predict single stone penetration in very soft clay. This model is developed based on existing static bearing capacity formulae and validated using laboratory experiments. The results agree well with the theoretical formulation. Field tests on board of FFPV Stornes are performed in the northern part of the North Sea to define all relevant processes and corresponding magnitude. Based on calculations performed seabed deformation due to consolidation forms the largest individual factor of volume loss. With one fifth of the complete loss single stone penetration embodies another important factor. However its influence is smaller than initially expected. Multi stone penetration causes the particles to penetrate even deeper into the soft soil. The stone penetration processes combined represent almost half of the average expected volume loss. Based on the validated single stone model and the described supplementary processes the total volume loss is modelled. A fit between the field results and the modelled penetration is determined to check if all processes are described and to validate their magnitude. The match between the measured and calculated processes suggests that all the right processes are taken into account. However to be able to predict the volume loss, accurate derivation of each process will have to be optimised. So far no clear conclusion can be drawn. The measurement accuracy currently used to obtain height deficits lacks the precision to undoubtedly state the influence of all separate processes. Important factors such as seabed erosion and the presence of sludge on top of the seabed, let alone the inaccuracy of the provided low stress soil properties, cause a lot of uncertainty in the determination of the individual process influence.