Ash dieback (ADB) has become a major problem across European ash forests. The disease is caused by the invasive pathogenic fungus Hymenoscyphus fraxineus that results in sever crown damage, and especially in young ash trees often rapid mortality. Fortunately, several studies have found that a fraction of F. excelsior trees remain healthy with only few symptoms. Heritability and additive genetic variance in tolerance have been estimated from progeny trials to be moderate to high, with limited genotype x environment (GxE) interaction. High levels of infection pressure are often present in forest areas with newly planted trees revealing their level of tolerance within few years. These features make ADB tolerance suitable for phenotypic selection and led to the hypothesis that strong forward phenotypic selection can create a new generation of ash trees with substantial reduced susceptibility to the disease. As a coordinated effort, this approach is tested by establishing selection trials in Norway, Sweden, Denmark and Lithuania. The joint efforts include sharing of genetic material to estimate GxE, but the program is based on development of tolerant ash trees from local genetic resources. Here we present the concept, challenges, activities and some preliminary results from the field trials.