Grassroots citizen science is a rapidly-expanding form of public engagement with science, which creates new opportunities and challenges for science, technology, and innovation governance. But what are these opportunities and challenges? How do they play out? And what do we learn from them? This paper seeks to provide empirically-grounded responses to these questions based on a comparative analysis of environmental grassroots citizen science initiatives in radiation pollution (Japan) and air pollution (Flanders, Belgium). Drawing on concepts from Science and Technology Studies, it casts grassroots citizen science as a technology of governance that is imbued with risks (e.g., risky data and devices, risky concepts and discourses) and risk taking (e.g., risky exchanges and relationships between citizens, scientists, policymakers, and other stakeholders). How we appreciate these risks is of key importance to the effective and democratic governance of environmental issues; yet, this question is rarely explicitly considered and addressed in citizen science processes and in citizen science literatures. The paper recommends foregrounding the assumptions and expectations that inform various risk conceptions, as well as how risks are handled and negotiated between concerned parties. Opening a space for these considerations can help parties to familiarize themselves with new (and old) risks and risk approaches, in ways that usefully unsettle (rather than disrupt) the science-technology-society nexus.