Subsurface Arsenic Removal (SAR) is a technique used for in-situ removal of naturally occurring arsenic in groundwater. This new technology was deployed recently on an experimental basis in two sites in rural Bangladesh, to address the pressing problem of rural drinking water supplies contaminated by arsenic. This article assesses whether and to what extent these first field experiments with SAR can be conceptualized as “socio-technical experiments” designed to incubate and improve radical technological innovations by serving as ‘living lab” “window” and/or “agent of change”. As per writings in transition theory, an experiment functions as a living lab if it permits testing, learning and improving upon a technological innovation. It functions as a window if it is able to facilitate communication and conversation by raising actors’ interest and enrolling new actors. It functions as an agent of change if it can successfully stimulate changes in potential users’ practices and behaviours. Through studying two SAR experiments, this article finds that this novel technology served as a living lab and window, but not (yet) as agent of change, partly because integrating social considerations (such as community buy-in, appropriate site selection and post-installation support) into SAR prototype design during field experimentation proved very difficult. A key obstacle was that the technical efficacy of the technology remained a primary concern during experimentation, and it was unsafe to make water deriving from experimental SAR units available to users. The technology thus remained an abstract idea and provided unable to stimulate behavioural changes amongst users. We conclude that there is a need to identify conditions under which real world experiments can serve as agents of change to facilitate sustainable uptake of arsenic safe technologies in rural developing country contexts.@en