The chemical characterisation of archaeological glass allows the discrimination between different glass groups and the identification of raw materials and technological traditions of their production. Several lines of evidence point towards the large-scale production of first millennium CE glass in a limited number of glass making factories from a mixture of Egyptian mineral soda and a locally available silica source. Fundamental changes in the manufacturing processes occurred from the eight/ninth century CE onwards, when Egyptian mineral soda was gradually replaced by soda-rich plant ash in Egypt as well as the Islamic Middle East. In order to elucidate the supply and consumption of glass during this transitional period, 31 glass samples from the assemblage found at Pergamon (Turkey) that date to the fourth to fourteenth centuries CE were analysed by electron microprobe analysis (EPMA) and by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). The statistical evaluation of the data revealed that the Byzantine glasses from Pergamon represent at least three different glass production technologies, one of which had not previously been recognised in the glass making traditions of the Mediterranean. While the chemical characteristics of the late antique and early medieval fragments confirm the current model of glass production and distribution at the time, the elemental make-up of the majority of the eighth- to fourteenth-century glasses from Pergamon indicate the existence of a late Byzantine glass type that is characterised by high alumina levels. Judging from the trace element patterns and elevated boron and lithium concentrations, these glasses were produced with a mineral soda different to the Egyptian natron from the Wadi Natrun, suggesting a possible regional Byzantine primary glass production in Asia Minor.