In this paper, we present the current practice of investigations of seafloor instabilities and deformation processes, based on extensive research conducted over the last years, which sets the scene for future research activities in this field. The mapping of the continental margins and coastal areas with ever increasing resolution systematically reveals evidence of instabilities and deformation processes, both active and palaeo-features. In order to properly assess the hazards and risks related to these features, an integrated and multi-disciplinary approach is essential, but challenging. Such an approach consists of combining field data (geophysics, geology, sedimentology, geochemistry and geotechnical data) with numerical simulations constrained by results from laboratory data. As such, it is of paramount importance to build a common knowledge base and understanding that unify these disciplines into more complete and conceptual models constrained by all the data. We review the status of this integrated approach adopted to palaeo-landslides (e.g., Storegga, Ana, Vesterålen) and recent deformations (Finneidfjord, Nice, Gulf of Guinea), allowing to identify gaps in our knowledge at these sites. By reviewing these case studies, one can conclude that each case remains highly site-specific in which both the regional and local geological-tectonic setting has a distinct effect of the type of instability or deformation taking place (or that can take place). Our knowledge on the actual triggers remains poorly constrained, and there is even ambiguity for historic landslides (e.g., Finneidfjord). Also our knowledge of the preconditioning factors is incomplete. There is a general lack of geotechnical data, both in situ and from laboratory, and therefore, modelling the dynamics (e.g., rheology) of the instabilities relies on a number of assumptions rather than facts. In addition, excess pore pressure and its evolution is one of the key parameters driving instabilities. Despite this fact, in situ (excess) pore pressure is rarely measured or monitored. Much work remains to be done to relate and integrate geotechnical data with geophysics, e.g., through inversion and rock physical models, in order to obtain additional quantitative information from the sub-surface, but also with respect to partial saturation (free gas, hydrate) and pore pressure behaviour, or lithologies. It is of critical importance to be able to identify the different processes which can lead to hazardous situations which includes establishing recurrence intervals (timing and frequencies, through event recognition and age control) and magnitudes, so that proper mitigation measures can be developed. In this perspective, the smaller-scale instabilities deserve much attention, as there are many instances where such features had far-reaching consequences for society (e.g., Nice, Finneidfjord). In that perspective, human interferences (e.g., exploitation, drilling, blasting, loading) must be one of the factors taken into consideration.