The burst pattern of electrical activity recorded from pancreatic B-cells in response to 11 mM glucose shows a large islet to islet variability. The relationship between burst frequency and glucose sensing (the threshold for electrical activity and the graded increase in electrical response to glucose, i.e. active phase %) has not been investigated within the same islet. In this work, we show that low HCO3 (5 mM) Hepes buffered solutions reversibly reduce the frequency of bursts compared to control (25 mM) HCO3 buffered solutions in the same islet. There was no change in the threshold or active phase (%). Using the mathematical model of Sherman et al. 1988, we explored mechanisms for a change in frequency independent of a change in active phase (%). Increased exchangeable calcium pool size and increased cell to cell coupling were the two theoretical treatments which could reproduce the experimental data. We conclude that burst frequency can be modulated independent of the active phase and that alteration of a calcium pool size best fits the experimental data.