ABSTRACT Climate variability and change can have important impacts for crop production. Therefore, the aim of this study is to investigate projections of the wheat yield in an increasingly warm climate. To address our objectives, we determined relationships between wheat yield in Spain and large‐scale variables. Partial least squares regression was applied to determine the modes of the climate variables that drive wheat‐yield variability, revealing a significant influence of surface solar radiation. Based on seasonal patterns of solar radiation, we determine models to estimate inter‐annual wheat‐yield variability. We find that the performance of the models based on solar radiation is better than that of earlier studies based on temperatures and precipitation variables. In this way, we use simulations of the Coupled Model Intercomparison Project Phase 5 (CMIP5) to project wheat‐yield trend under warming climate by implementing direct statistical downscaling. The expected range of projected wheat yield trend for 21st century indicates decreases of about 6–8% across Spain. The suggested models could be applied for adaptation and planning.