In the light of current global climate change, gene introgression from wild relatives is a key strategy for increasing adaptation of crops that lost their variation during evolutionary bottleneck. Fall‐sown safflower leads to earlier spring growth and potentially higher productivity than the spring‐sown safflower, because of escaping from the terminal stresses of drought, heat, disease, and pests as well as efficient use of winter precipitation. During the last decade, we developed advanced generations of three sets of recombinant inbred lines (RILs) derived from three crosses of Carthamus tinctorius × Carthamus palaestinus (TP), C. tinctorius × Carthamus oxyacanthus (TO) and C. palaestinus × C. oxyacanthus (PO). The objectives of this study were to assess the efficiency of gene introgression from two wild relatives of safflower and genetic variation of derived RILs in terms of adaptation to autumn sowing and cold tolerance in the field. The results showed a continuous variation for seed yield and oil content in RILs, indicating the existence of transgressive segregation in lines. Comparison of three interspecific populations indicated that lines derived from TP crosses showed high potential for autumn planting, winter hardness, and productivity of seed and oil yield. However, the RILs derived from PO and TO had higher chlorophyll fluorescence (Fv/Fm) than TP population. Our results indicated that restoring lost genes in safflower is possible by introgression from its close relatives (especially C. palaestinus species). Identified superior RILs can be checked for other biotic and abiotic tolerance to develop new varieties in this valuable oil seed crop.