Often, tsunami “sources” have been treated as a quasistatic problem. Initial studies have demonstrated that, for earthquake rupture velocities in the span of , the kinematic and static part of the tsunami can be treated separately. However, very slow earthquake rupture velocities in the span of have not been included in tsunami analytical or numerical modeling. Here, we calculated the tsunami efficiency, extending Kajiura’s definition for different models. We demonstrated that rupture velocity cannot be neglected for very slow events, that is, rupture velocities slower than . We also examined the relation of magnitude, earthquake rupture velocity, and tsunami amplitude to the efficiency of very slow tsunamigenic earthquakes. Hypothetical megathrust earthquakes () with very slow rupture velocities amplify energy from 10 to 60 times larger than moderate to large earthquakes () in the direction of rupture propagation.