It is usually concluded from laboratory tests that deicing salt scaling resistance of High-Volume Fly Ash concrete (HVFA concrete) is less than satisfactory. This is mainly attributed to the partial adsorption of the applied air entraining agent (AEA) by the unburnt carbon present in the fly ash. As a consequence, the presence of an adequate air void system in HVFA concrete can not always be garantueed. This explains why the use of HVFA concrete in pavements exposed to freezing and thawing with deicing salts, is still not generally accepted. Therefore, we carefully evaluated the air void system of this concrete type in the hardened state. This paper presents a full assessment of this air void system for two HVFA compositions with a total binder content of 450 kg/m³ (50 % cement, 50 % fly ash) and a water-to-binder ratio of 0.35. The two mixtures mutually differed in the type of fly ash used. For both of them, the minimum required air content in the fresh state was achieved with an AEA dosage of 5.0 ml/kg binder. The air void system was evaluated qualitatively using microscopic analysis on thin sections. The air content (≥ 4 %) and spacing factor (≤ 200 μm) were quantified with the Rapidair 457 apparatus. Although these two properties were similar for the two mixtures and they both met the applicable criteria, the air void system still seemed a bit less pronounced under the microscope for one of the mixtures. The salt scaling resistance of the latter mixture was also found to be significantly less after 28 days of curing and 28 severe freeze/thaw cycles. The mass loss due to salt scaling was found to be equal to more or less 1 kg/m², the maximum allowed value according to the applicable European standard, while the amount of scaled off material for the other mixture was much less (0.44 kg/m²).