Sunflower ( Helianthus annuus L.) breeding programs typically target heterogeneous regions where large genotype × environment interactions (GEI) complicate genetic progress. Some understanding of the underlying factors nature and repeatability of GEIs can help to accommodate their effects. This review summarizes the findings of a series of studies conducted in Argentina with the goals of understanding the effect of GEIs on sunflower yield progress and design breeding strategies aimed at impacting on future rates of genetic advance. Pattern analyses revealed that the central and northern subregions are different mega-environments (ME) which suggests that yield gains in the northern ME would have been reduced if selection had been done in central environments and vice versa. Photoperiod and minimum temperature during the reproductive phases are the main environmental factors underlying the observed GEIs. Genetic gain analyses indicated clear and continuous improvements in oil yield due to plant breeding during 25 years and confirmed the prediction about lack of yield gains under the northern ME from selection done in central environments. The central ME showed the highest and most consistent genetic progress probably because it concentrates most breeding programs and shows the highest repeatability. Breeding progressed to merge the best combinations of two original hybrid groups reducing the original maturity range within MEs and concentrating on different maturity types among MEs. The model of correlated response to selection indicated that dividing the central- north region of Argentina in central and northern MEs and exploiting specific adaptation would result in faster genetic progress. In the central ME there is scope to replace years with locations at no costs in heritability. In the northern subregion trials should be conducted over at least four years. Late planting dates in a central location associated positively with the northern environments in terms of genotypic discrimination; which constitutes an opportunity for indirect selection for the northern ME. Physiological models served to establish the importance of attributes and processes affecting crop performance in the post-anthesis phase in determining the observed genotypeby- planting time interaction. The diameter of the empty center and canopy stay green were found to be efficient secondary indicators related to seed set and duration of grain filling respectively - indicators that can be used to improve the efficiency of selection for specific adaptation to late plantings and to the northern ME.