Nowadays the most important sectors in the strategy for the future of the port of Rotterdam are container handling, chemicals and distribution. To offer these sectors the opportunity to grow and renew, space is needed. By means of the construction of Maasvlakte II this space can be given. This land reclamation is planned to be located in the area between the Euro-Maasgeul in the north and the current Maasvlakte in the west and the extended demarcation line in the south. The planned extension will be done in two phases. The first one includes 700 ha with a length of breakwater of 2.7 km, whose construction is planned to start in 2006/07 and the second phase, which includes 300 ha more, with a length of breakwater of 1.3 km is planned to start in 2013/2023. In this study the cross-section of the breakwater, which protects the new area of the land reclamation, is analysed. Classical deterministic design could provide a preliminary geometry for the breakwater, but the dimensions of it are too big and also the costs. Therefore a probabilistic optimization could be made in order to check if a reduction or growth, in the geometry, can provide an economical optimum geometry with a substantial save. First a classical deterministic design is made. The most important elements of the crosssection are determined with the classical formulas and design guidelines. The following elements are analysed: - The armour layer - The toe - The secondary armour - The core - The filter system to establish the supporting bottom material - The crest height When the dimensions are given for all the elements, the geometry of the breakwater is established for the deterministic design. Afterwards the construction costs are determined for the breakwater solution. The deterministic design results in an element weight of 18.8 tons (6.6 m3) and crest height of NAP+18 m. Economic consequences of the different failure mechanisms are not taken into account. The crest height is normally dependent on the construction method. In that case, the construction method does not produce a sensitive reduction in the breakwater geometry because the security level required in the determination of the crest height is too restrictive. After the deterministic design, the probabilistic optimization takes place..