As part of an ongoing program to constrain the factors affecting phase stability in nuclear waste containment glasses, the thermodynamic and kinetic factors influencing molybdenum oxide solubility in a borosilicate melt have been studied. A simplified system has been chosen to study how Mo oxide is transferred from solid effluent residues (calcines) to glass. The model calcines used are pure Na2MoO4 and a more complex mixture of Na2MoO4 and Al-bearing phases. These solid residues are mixed with a borosilicate glass frit and heat treated isothermally at temperatures in the range 700–1000 °C for times ranging from 2 min to 8 h. The mineralogy of the mixtures has been followed as a function of temperature and time using XRD and 27Al and 23Na MAS NMR spectroscopy, while the structure and composition of the amorphous phase have been studied by 11B MAS NMR and electron microprobe. The results obtained illustrate that while the dissolution of Al-rich calcines can modify the structure of the borosilicate network this does not affect either the MoO3 content of the liquid at saturation, or the characteristic time-scale required to reach saturation. In detail, the variation of the Mo content of the liquid at saturation can be described by an Arrhenian temperature dependence with an activation energy of 30 ± 5 kJ/mol and the time-scale required to reach saturation, can be described by an exponential function of time. This time-scale also shows Arrhenian temperature dependence with an activation energy of 225 ± 25 kJ/mol, a value identical within uncertainty to that associated with viscous flow. Using the experimental results, a numerical model for dissolution along variable temperature-time paths has been developed. These models indicate that for systems that are capable of complete dissolution grain-size is one of the most important parameter controlling the efficiency of the dissolution process.