The increased incidence of extreme temperature events due to global climate change poses a major challenge for crop production. Ability to increase temperature tolerance through genetic improvement requires understanding of how crops and their wild relatives respond to extreme temperatures. We developed a high-throughput technique to evaluate tolerance to freezing stress (FS) and heat stress (HS) in wild, crop–wild hybrid and cultivated sunflower (Helianthus annuus L.). We also investigated whether trade-offs exist between stress tolerance and growth under benign conditions. Eleven experiments were performed under a combination of growth-chamber and field conditions. In growth-chamber experiments, FS and HS consisted of exposing acclimated plants at the 2–4-leaf stage to temperatures ranging from to –2.5°C to –4°C for 2–4 h and from 52°C to 54°C for 2–3 h. In the field, plants were grown for 32 days during midwinter (FS: average T_mean = 9.9°C and T_min = 3.8°C) or for 10 days in a heat tent (HS: average T_mean = 30.1°C and T_max = 43.3°C). We observed large differences in tolerance to FS and HS between wild and cultivated sunflower. Wild sunflower showed higher FS tolerance than cultivated in both growth-chamber and field experiments, whereas cultivated sunflower showed higher HS tolerance in growth-chamber experiments. No differences in HS tolerance were observed in the field. Crop–wild hybrids generally showed intermediate HS and FS tolerance. We found no evidence of a growth-tolerance trade-off, which suggests that tolerance might be introgressed into elite germplasm without growth penalties. The study reveals that wide genetic variation for the tolerance to extreme temperatures exists in the primary gene pool of sunflower.