Freezing stress resistance is composed of several components namely tolerance, avoidance and acclimation ability. These three components of freezing stress are heritable traits. We have demonstrated that progress in the improvement of freezing stress resistance can be made by individually selecting for various components of this resistance and then recombining them to get the desired plan. Freeze-thaw injury in carrots is manifested as damage to the foliage, cracks on the roots (especially on the crown), and crown root rot. We found that foliage damage following freeze-thaw stress was related to the tolerance of the foliage to ice formation. The formation of cracks in the crown and root tissue was related to formation of ice itself. The carrot breeding lines we tested varied considerably for the crown position in relation to soil surface. The carrot crowns and roots below the soil surface will be better in avoiding ice in the tissue, thus avoiding cracks. The freeze-thaw injury observed on the foliage in the field was highly correlated to the freeze-thaw tolerance of leaf tissue (measured as ion leakage from the leaf tissue) determined by controlled freeze-thaw test in the laboratory. Based on this work we developed a breeding strategy to improve frost hardiness in carrots by combining the characteristics that avoid ice in the crown and root tissues (e.g., crown position underground) with the characteristics that reduce foliage and root injury by ice (freezing tolerance of foliage). By using this strategy we were able to successfully obtain the desired plant. Two hardy carrot hybrids (Eskimo, Artico) were released by Vilmorin and their hardy characteristics have been confirmed under field conditions.