The critical step for any subsequent instrumental analysis of cementitious binders is to stop their hydration reactions, i.e., to remove free water. One of the most available techniques is a solvent exchange method. However, the solvents are known to be strongly bound in ordinary Portland cement (OPC) paste and alter the results of thermogravimetric analysis (TGA) and sensitive hydrates, while their effect on TGA response of alkali-activated slag (AAS) has not been comprehensively investigated. Therefore, the objective of this paper is to track the effects of fundamental aspects of the solvent exchange on the TGA response of AAS with different sodium activators (hydroxide, carbonate, waterglass) and to support these results by X-ray diffraction and effluent gas analysis. All solvents used (acetone, diethyl ether, isopropyl alcohol, ethanol, and methanol) affected the TGA response of all tested pastes, and their effect was enhanced by prolonged immersion time. All solvents induced an additional mass loss at around 800 °C and, especially for OPC paste, increased in situ carbonation, even in an inert atmosphere. Methanol and ethanol had a detrimental effect on ettringite and decreased the basal distance of the C–(A)–S–H gel, while they only marginally affected gaylussite. For AAS, hydration stoppage by washing out the alkali-rich pore solution with water was also investigated and can usually be recommended (except for its detrimental effect on gaylussite), as it is more efficient than organic solvents, which lack solubility for activators. Methanol and ethanol are the most suitable alternatives, particularly for NaOH.