CO2 emissions from cement production currently represent around 6% of global CO2 emissions. However, cement concrete absorbs CO2 from the atmosphere because of carbonation (i.e., penetration of atmospheric CO2 inside bulk concrete). Carbonation has beneficial effects on the mechanical resistance of cement concrete. However, carbonation also has adverse effects because it provokes a decrease in pH that favors later corrosion of reinforcing bars and thus reduces service life. Current European standards provide recommendations concerning reinforcing concrete covers, but these are not based on actual service-life durations. Thanks to a previously developed carbonation model combined with sensitivity analysis and LCA, we compare Climate Change indicators of 1 m2 of reinforced concrete cover over a 100-years service life exposed to XC4 conditions in Madrid, obtained on one hand by using current standards and on the other hand with concrete-cover depths calculated with our carbonation model. Our results show that cement strength class is a key parameter to both increase durability and decrease climate-change impacts. When the carbonation model is used to optimize both durability and climate-change impacts, it drives to considerable and significant improvements. Finally, climate-change indicators predicted from our carbonation model are not linearly linked to carbon intensity of cements, which is a current argument of so-called green cements. The values of indicators presented in this article cannot be generalized: They mainly depend on the geographical location. However, the model and key action levers are general. Using high cement strength classes and low water-to-cement ratios allows use of lower concrete-cover depths and thus save amounts of concrete compared to the standard. This generates an important benefit in terms of climate-change impacts for identical service lives and improved mechanical resistance. Thus, considering the huge impact of cement and construction industry on climate change, we plead for a revision of standards which, instead of thresholds based on simplified models, should provide certified tools enabling the best design for every situation. This article met the requirements for a gold/gold JIE data openness badge described at http://jie.click/badges.