Beryllium wastes are produced by nuclear industry. One way to manage them is their encapsulation in cements. The main risk of this conditioning is the aqueous corrosion, which leads to the hydrogen production and cracks causing a loss of radioactivity confinement. The corrosion can be limited by the formation of the hydroxide solid phase Be(OH)2(s). The stability domain of this phase was calculated in water as a function of the pH: M. Pourbaix has calculated a stability domain from 2.9 to 11.7 for a 10⁻⁴ M beryllium concentration, while according to our calculation with more recent thermodynamic data, it is stable from 5.3 to 13.5. Based on Pourbaix results, beryllium cannot be conditioned in the mainly used cement for nuclear waste, Portland cement, while it is possible according to our calculations. Experimental measurements were achieved to select the data set most in agreement with the experimental observations. The beryllium reactivity has been examined in matrices having different pH pore solution: brushite cement (pH 1.75–6.44), magnesium phosphate cement (pH 5.6–8.4), calcium-sulfoaluminate cement (pH 10.9–12.3), Portland cement (pH 12.5–12.9) cements and activated slag (pH 12.9–13.8), by measuring the open circuit potential and by electrochemical impedance spectroscopy. The experimental results agree with the more recent thermodynamic data. Beryllium corrosion is too high in the brushite cement, leading to a high hydrogen production. This matrix can then not be envisaged for the conditioning of Be waste. If the beryllium is encapsulated in the activated slag, the highly alkalinity is too high in the early age, leading to a high aqueous corrosion. Activated slag are also not suitable for Be conditioning. The main conclusion of this paper is that beryllium can be encapsulated in safe conditions in Portland, magnesium phosphate and calcium sulfoaluminate cements.