Questions: Studies in functional biogeography have mostly focused on unmanaged ecosystems, and neglected testing how management intensity affects community-level response of plant traits to bioclimatic gradients. We hypothesize that trait–climate relationships for arable weeds spontaneously establishing in croplands subject to intensive management should differ from the relationships characterizing less intensively managed grassland ecosystems. Location: France. Methods: We computed community-weighted means (CWM) and variances (CWV) of 954 and 5,619 cropland and grassland plant assemblages, respectively, for three fundamental leaf traits (specific leaf area, SLA; leaf dry matter content, LDMC; leaf nitrogen content, LNC). Based on growing season length accounting for both temperature and soil water limitations (GSLtw), we compared trait–climate relationships between herbicide-free croplands and grasslands, and between herbicide-free and herbicide-sprayed cropland assemblages. The contribution of beta-diversity to the trait–climate relationships was then evaluated using multiple regression on distance matrices. Results: Distinct trait–climate relationships characterized herbicide-free cropland and grassland plant assemblages. CWM of all traits showed weaker relations with GSLtw in cropland relative to grassland assemblages. CWV of LDMC and LNC responded more sharply in croplands. Furthermore, no herbicide effect on trait–climate relationships was detected within cropland assemblages. These results seem to be explained by a greater taxonomic beta-diversity along the GSLtw gradient for grasslands. Conclusions: Specific trait–environment relationships characterize croplands, underlining that management intensity greatly affects trait–climate relationships for plant assemblages. Deciphering the interplay between land use intensification and climate is critical to accurately forecast functional vegetation changes in response to global changes, and hence to foster actions enhancing ecosystem resilience.