AIM: To test whether bioactive glasses kill microbiota via mineralization or the release of ions other than sodium. METHODOLOGY: Flame-spray synthesis was applied to produce nanometric glasses of different sodium content and constant Ca/P ratio: 28S5, 45S5 and 77S. Calcium hydroxide and nanometric tricalcium phosphate (TCP) were used as controls. Apatite induction was monitored by Raman spectroscopy. Bovine dentine disks with adherent Enterococcus faecalis cells were exposed to test and control suspensions or buffered solutions for 1 h, 1 day and 1 week. Colony-forming units were counted and disks were inspected using scanning electron microscopy. Suspension supernatants and solutions were analysed for their pH, osmolarity, calcium and silicon content. RESULTS: Sodium containing glasses induced pH levels above 12, compared with less than pH 9 with sodium-free 77S. Calcium hydroxide, 45S5 and 28S5 killed all bacteria after 1 day and lysed them after 1 week. TCP caused the highest apatite induction and substantial calcification on bacteria adhering to dentine, but did not reduce viable counts. 77S achieved disinfection after 1 week without visible apatite formation, whilst the buffer solution at pH 9 caused only minimal reduction in counts. CONCLUSION: Bioactive glasses have a directly and an indirectly pH-related antibacterial effect. The effect not directly linked to pH is because of ion release rather than mineralization.