To reduce the impact of the embodied energy and of the carbon dioxide emissions of concrete, the use of supplementary cementitious materials (SCMs) in the cement industry has become a common practice. However, the practical experience on such cements is limited due to the lower strength at early ages and the concern about long term performance, especially potentially higher carbonation rates due to a lower capacity to bind CO2. To increase the protection of the steel in the concrete where carbonation might be a risk, the full understanding of the microstructure changes due to carbonation is necessary. The main goal of this thesis is to characterise and to quantify the correlation between transport properties and modification of the microstructure in the carbonated concrete. We propose a new approach to obtain a representative specimen of “fully carbonated” material in relatively short time. The innovative idea of using a thin cement paste sample allowed the characterization of the gas diffusion properties in naturally carbonated cement paste. This, together with the measured changes in phase assemblage and pore structure, can be used to better understand the reactive transport model. The adaptation to natural exposure conditions ensures obtaining representative results. Investigating the governing parameters of carbonation and emphasizing the influence of the non-carbonated reference advanced the understanding of the carbonation mechanism, especially in the low carbon binders. This study sheds a new light at the problem of overestimating the effect of carbonation on the microstructure of the cementitious material. In particular, it shows the importance of subtracting the effect of the drying taking place during carbonation, to diminish the risk of attributing the changes in porosity only to carbonation. The monitoring of the changes in the phase assemblage and the assessment of the carbonation coefficient during exposure to carbonation showed a promising performance of the new ternary blend with 50 % clinker replacement by burnt oil shale, slag, and limestone. The obtained results on characterisation of the carbonating binders can be used to improve carbonation models, that are essential to predict the resistance of new types of cements. That could help to assess the balance between the benefits of using alternative materials, and the potential danger of low resistance against carbonation. The profit will be to make cement and concrete even more sustainable and a more environmentally friendly construction material.