To investigate the dynamics of submersible mussel rafts, the finite element program Aqua-FETM, developed by the University of New Hampshire(UNH), was applied to rafts moored at the surface and submerged. The submerged configuration is used to reduce wave forcing and to avoid contact with floating ice during winters in northern waters. Each raft consists of three pontoons connected by a grid framework. Rafts are intended to support densely spaced mussel ropes hung from the framework. When submerged, the pontoons are flooded, and the raft is held vertically by floats attached by lines. The computer models were developed in Aqua-FE? to simulate the effects of waves and current. They were validated by comparison with wave tank results by use of a 1/10 scale raft physical model. Comparisons showed good agreement for the important heave(vertical) and pitch(rotational) motions, though there was a tendency towards conservative results for wave and current drag. Full-scale simulations of surface and submerged single raft and two rafts connected in tandem were performed. Submerged raft wave response was found to be reduced relative to that at the surface for both the single and two-raft configurations. In particular, the vertical motion of mussel rope connection points was significantly reduced by submergence, resulting in reduced potential for mussel drop-off. For example, the maximum vertical velocities of mussel rope attachment points in the submerged two raft case were 7%-20% of the corresponding velocities when at the surface. To investigate the dynamics of submersible mussel rafts, the finite element program Aqua-FETM, developed by the University of New Hampshire (UNH), was applied to rafts moored at the surface and submerged. The submerged configuration is used to reduce waves forcing and to avoid contact with floating ice during winters in northern waters. Each raft consists of three pontoons connected by a grid framework. Rafts are intended to support densely spaced mussel ropes hung from the framework. When submerged, the pontoons are flooded, and the raft is held vertically by floats attached by lines. The computer models were developed in Aqua-FE? to simulate the effects of waves and current. They were validated by comparison with wave tank results by a 1/10 scale raft physical model. Comparisons showed good agreement for the important heave (vertical) and pitch (rotational) motions, though there was a tendency for conservative results for wave and current drag. Full-scale simulations of surfac e and submerged single raft and two rafts connected in tandem were performed. Submerged raft wave response was found to be reduced relative to that at at the surface for both the single and two-raft configurations. In particular, the vertical motion of mussel rope connection points was significantly reduced by submergence, resulting in reduced potential for mussel drop-off. For example, the maximum vertical velocities of mussel rope attachment points in the submerged two raft case were 7% -20% of the corresponding velocities when at the surface.