Industrial cropping systems are increasingly simplified, with fewer crops being grown, requiring extensive use of inputs, and contributing to environmental pollution and climate change. Solutions are needed to reduce the negative effects, while retaining productivity in the face of increasingly frequent detrimental climatic conditions. Crop rotational diversity has shown promises to increase staple cereal yields, especially under low fertilisation regimes and years of low productivity. However, it is unclear how different levels and types of crop rotational diversity mediate the interaction between productivity, contrasting levels of fertilisation, and explicit climatic conditions. Moreover, crop diversification substitutes staple cereals with alternative crop types, e.g., legumes, broadleaves, and perennial mixtures of grasses and legumes, raising the question whether sufficient and sufficiently nutritious food can be produced with less nitrogen inputs in diverse rotations. Using data from 32 long-term field experiments (10-64 years) from Europe and North America, I show that crop rotational diversity, expressed as species diversity or functional richness, enhances cereal yields over time, particularly under regimes of low fertilisation. I also show that crop rotational diversity can reduce cereal yield losses caused by detrimental climatic conditions. Using a sub-set of the European data, I also show that increasing functional richness can increase the outputs of human-available calories and macronutrients, i.e., carbohydrates, proteins, and fats, with increasing benefits over time, and requiring less nitrogen than cereal-only rotations. In summary, crop rotational diversity can benefit food security and sustainability, and provide climate adaptation to cropping systems.