Faba bean is a legume crop with high protein content and considerable potential for wider cultivation in cool climates. However, it has a reputation for having unstable yield with large interannual variability, mostly attributed to yearly variation in rainfall. In this study, 17 commercial cultivars of faba bean were evaluated for seed yield, yield stability and the relationship between seed yield and protein content at four locations in Denmark and Finland during 2016–2018. We found that location and year effects accounted for 89% of the total seed yield variation. Cultivar × environment (GxE) interactions were small (2.4%) and did not cause reranking of cultivars across environments. Yield stability contributed little to the mean yield of the cultivars, as high‐yielding cultivars consistently outperformed the lower yielding genotypes, even under the most adverse conditions. Similarly, GxE effects on protein content were limited, and we found an overall negative correlation of −0.61 between seed yield and protein content for the cultivars and environments studied. These data may be helpful for selecting cultivars for field use or for use in breeding programmes, considering that future faba bean pricing could depend on both protein quantity and concentration.

Faba bean is a legume crop with high protein content and large potential for cultivation in the Northern latitudes. However, it has a reputation for being an unstable crop with large interannual variability, mostly explained by yearly variation in rainfall. Consequently, the objective is to breed cultivars with high seed yield and high yield stability. In this study, 17 commercial cultivars of faba bean were evaluated for seed yield, yield stability and trade-off between seed yield and protein content in four locations in Denmark and Finland during 2016-2018. We found that location and year effects accounted for 72% of the total seed yield variation. Cultivar by environment interactions (GxE) were found to be small and did not cause re-ranking of cultivars in different environments. Yield stability contributed little to the mean yield of the cultivars because high-yielding cultivars consistently outperformed the lower yielding genotypes, even under the most adverse conditions. The latter was also the case for total protein yield quantified as total yield multiplied by seed protein percentage. Although we found a strong negative correlation of -0.64 between yield and protein content, a few cultivars produced high yields while maintaining a relatively high protein content, suggesting that these traits may to some degree be genetically separable.