The influence of electron irradiation on the glass transition temperature, Tg, of poly(ether ether ketone) (PEEK) has been investigated by differential scanning calorimetry, up to a dose of 100 MGy. For amorphous PEEK, the observed Tg increases linearly with absorbed dose at a rate of 0.18 degreesC MGy(-1). This indicates the formation of crosslinks, as deduced elsewhere. Above similar to 50 MGy, these crosslinks prevent crystallization on heating to above Tg, whereas at lower doses, the polymer is able to crystallize to some degree. Qualitatively, the variations in Tg seen in these partially crystalline samples can be explained in terms of a number of factors. The crystals serve to constrain the amorphous fraction, resulting in a direct elevation in Tg. This is reinforced through the rejection of crosslinks from crystalline regions, so resulting in an increase in the local amorphous crosslink density. Conversely, crosslinking can also serve to inhibit crystallization. Quantitatively, models based solely upon the percentage crystallinity are, however, unable to account fully for the variations in Tg that are seen. In contrast, a model based on both overall crystallinity and lamellar thickness gives good agreement with experiment over the complete crystallinity range.