The Pacific coast of Colombia has witnessed several tsunami events in the last century. However, the physical damage to structures at the time was not recorded, even when these events have brought overwhelming destruction to coastal communities. This lack of data increases uncertainties when assessing future tsunami impacts. Nonetheless, other regions of the world, which have been also affected by tsunami, have collected damage information that has been used to build empirical tsunami fragility functions. Based on a demand parameter, such as the tsunami inundation depth, a fragility function expresses the probability of building damage. Hence, regions with no recent tsunami damage cannot develop empirical fragility functions since post-disaster damage survey data is not available. One alternative for these regions is to develop analytical fragility functions out of computational nonlinear structural analysis. Reinforced concrete buildings and palafitte houses are selected to perform a comparative analysis of collapse damage assessment expressed in empirical and analytical fragility functions.