Following several decades of research, there is not yet a convincing vaccine against shigellosis. It is still difficult, in spite of the breadth of strategies (i.e. live attenuated oral, killed oral, subunit parenteral) to select an optimal option. Two approaches are clearly emerging: (i) live attenuated deletion mutants based on rational selection of genes that are key in the pathogenic process, and (ii) conjugated detoxified polysaccharide parenteral vaccines, or more recently conjugated synthetic carbohydrates. Some of these approaches have already undergone phase I and II clinical trials with promising results, but important issues have also emerged, particularly the discrepancy between colonization and immunogenic potential of live attenuated vaccine candidates depending upon the population concerned (i.e. non endemic vs. endemic areas). Efforts are needed to definitely establish the proof of concept of these approaches, and thus the need for clinical trials which should also soon explore the possibility to associate different serotypes, in response to serotype specific protection against shigellosis. More basic research is also required to improve what we can still consider as first-generation vaccines, and to explore possible new paradigms including the search for cross-protective antigens.