It has been increasingly recognised that calculation of the disease burden due to populations, such as in Bangladesh, extensively using hazardous arsenic bearing well waters, must explicitly account for the trade-off between diarrhoeal disease incidence and that of arsenic-related diseases. This is because it is likely that moves to alternative drinking water sources, be they surface waters or even more distant groundwaters, without further mitigation would result in a concurrent increase in diarrhoeal disease. In this paper, we update the model of Lokuge of the effects of such arsenic mitigation on disease burden in Bangladesh, using updated population data and background disease estimates. We run a critical pathway analysis on the model using Standardised Mortality Ratios (SMRs) for diabetes mellitus and ischemic heart disease from different epidemiological studies recently reviewed by Navas-Acien. Our analysis agrees with that of Lokuge that mitigation simply involving the substitution of a range of surface waters for well water sources with As > 50 microg/L would have a net positive impact on disease burden, as determined by deaths and Disability Life Adjusted Years (DALYs). In contrast, however, there is considerable ambiguity in the analogous results for mitigation for all the population exposed to well water with As > 10 microg/L. Depending upon the data source chosen for diabetes mellitus and ischaemic heart disease SMRs, such mitigation is modelled to have either a positive or a negative net impact on overall disease burden. The modelled negative impacts are entirely commensurate with the rationale for seeking groundwater as an alternative to surface waters as a drinking water supply, and highlight the practical requirement for multiple mitigation strategies, including those directed at ensuring the microbiological safety and continued protection of any alternative water supplies. Our study highlights the need for (i) adequate epidemiological studies involving multiple exposure categories, ideally resulting in an accurate dose-response relationship for arsenic uptake and the non-malignant high incidence conditions diabetes mellitus and ischemic heart disease for individuals with the socioeconomic and nutritional status of the Bangladeshi populations, and (ii) refined estimates of the diarrhoel disease burden arising from usage of surface waters.