The alteration of SON68 glass (inactive R7T7 type glass) in alkaline (NaOH and KOH) solution was studied at pH 11.4 at 90 °C and at an S/V ratio of 65 cm−1. Under these conditions the glass alteration rate initially diminished due to the formation of a protective gel layer, as observed during alteration of the same glass at lower pH (7–10). After a time varying from 14 to 28 days, depending on the nature of the base, alteration subsequently resumed. This study confirms the latest results showing that the phenomenon is due to the crystallization of zeolite phases such as analcime or merlinoite at the gel surface. These phases initially consume the aluminum in solution, then all the aluminum and a fraction of the silicon in the gel. When the solution pH is artificially decreased to 9.5 during the renewed alteration stage, the zeolites gradually dissolve and glass dissolution ceases. The instability of these secondary phases at low alkaline pH shows that nucleation of the phases is not the phenomenon limiting the glass alteration kinetics, and that a resumption of alteration due to the precipitation of these phases is unlikely in nuclear glass disposal conditions (pH 7–10).