Biological introductions of species to regions outside their known natural distribution are considered a major threat to native marine biodiversity and a key consideration for ecological management. For most invasive species in marine systems, however, little is known about potential impacts. If we are to increase our knowledge of the processes and mechanisms behind the spread of nonindigenous species or determine economic or ecological impacts, manipulative ecological field experiments are the best way to unambiguously ascribe causal relationships. For studies of invasions, such research may result in species spread and the establishment of new viable populations. Is it ethical then, to take the risk of potentially modifying or endangering other species, populations or ecosystems? Is it possible to mitigate the risks? Or should invasion ecologists work under restrictions that limit their ability to fully assess the impact of invaders? Consideration of the ethics of experimentation is rarely carried out. As a consequence, we propose a decision model that includes possible risk of escape/establishment versus the value of the research to allow researchers and/or managers to critically evaluate what type of experimental approach is appropriate.