abstract Climatic conditions and hence climate change influence agriculture. Most studies that addressed the vulnerability of agriculture to climate change have focused on potential impacts without considering adaptation. When adaptation strategies are considered, socio-economic conditions and farm management are often ignored, but these strongly influence current farm performance and are likely to also influence adaptation to future changes. This study analysed the adaptation of farmers and regions in the European Union to prevailing climatic conditions, climate change and climate variability in the last decades (1990–2003) in the context of other conditions and changes. We compared (1) responses in crop yields with responses in farmers’ income, (2) responses to spatial climate variability with responses to temporal climate variability, (3) farm level responses with regional level responses and (4) potential climate impacts (based on crop models) with actual climate impacts (based on farm accountancy data). Results indicated that impacts on crop yields cannot directly be translated to impacts on farmers’ income, as farmers adapt by changing crop rotations and inputs. Secondly, the impacts of climatic conditions on spatial variability in crop yields and farmers’ income, with generally lower yields in warmer climates, is different from the impacts of temporal variability in climate, for which more heterogeneous patterns are observed across regions in Europe. Thirdly, actual impacts of climate change and variability are largely dependent on farm characteristics (e.g. intensity, size, land use), which influence management and adaptation. To accurately understand impacts and adaptation, assessments should consider responses at different levels of organization. As different farm types adapt differently, a larger diversity in farm types reduces impacts of climate variability at regional level, but certain farm types may still be vulnerable. Lastly, we observed that management and adaptation can largely reduce the potential impacts of climate change and climate variability on crop yields and farmers’ income. We conclude that for reliable projections of the impacts of climate change on agriculture, adaptation should not be seen anymore as a last step in a vulnerability assessment, but as integrated part of the models used to simulate crop yields, farmers’ income and other indicators related to agricultural performance.