This study investigated the shear modulus degradation curves of reconstituted stabilized soil specimens and their evolution during curing up to 270 days. An experimental protocol is proposed to get as close as possible to the in situ curing conditions of a chemically (lime and hydraulic binder) and mechanically (compaction) improved soil placed in a high-rise geotechnical structure. Innovative solutions now make it possible to build high embankments (> 10 m) in stabilized soil. In the laboratory, a silty soil stabilized with 1% lime and 5% hydraulic binder, which is a classic treatment for linear projects, was cured with a pressure of 300 and 500 kPa, corresponding to a depth in the embankment of 15 and 25 m. These structures are expected to have high mechanical performance and must be able to withstand very small deformations (earthquakes, traffic loads). Experimental campaigns are performed in laboratory using torsion tests with a resonant column apparatus (RCA) to obtain the shear modulus G as a function of distortion. A comparison was made between normalized curing conditions (temperature of 20 °C and atmospheric pressure) and curing with confining pressure (temperature of 20 °C and 300 or 500 kPa pressure). For this aim, 74 torsional RCA tests were conducted on 16 compacted soil specimens. These showed that the curing pressure, applied just after compaction, brings the expected properties more quickly and confers superior properties of around 15% in the long term. A power law was proposed to estimate the Gmax of a stabilized silt specimen as a function of time, based on the Gmax measured during the first two days of curing. This empirical law gives the Gmax over the long term (270 days) for curing with or without confining pressure. Pioneering a combined curing-testing RCA approach, this study elucidates the short and long-term behavior of stabilized silts under realistic curing conditions.