The article addresses the theme of the validity of intervention-research in community psychology. There are three types of validity in this area: ecological validity, relational validity and transformational validity. Although all three types are important, relational validity and transformational validity are the most suitable for intervention-research in the community sphere, in the Authors’ opinion. Relational validity rests on processes of trust and mutual respect between researchers and participants as well as on building a relational context that makes possible physical and interpersonal access to the community. Transformational validity refers to processes of change promoted by the community thanks to the execution of the intervention-research and to the sustainability of the changes set in motion for the community.