More-than-human, multispecies, and animals’ geographic accounts of the city have tended to focus on large, charismatic, and wild organisms, to the detriment of spatially invisible other-than-humans central to urban reproduction. At the same time, urban microbial geographies have foregrounded embodied interactions between humans and microorganisms, whether they are symbiotic or pathogenic, often marginalising the material contributions of extracorporeal microbiomes to the urban fabric. Building from these two blindspots, this article focuses on microbial ecologies that live constitutively outside of (other-than-)human bodies and which are intimately caught up in the metabolic intensities and infrastructural environments of the urban realm. There are two key aims: 1) to explore different forms of urban microbial ecologies and 2) to examine their relationships with urban infrastructures and reproduction. My disciplinary lenses are animals’ geographies and urban ecology and my case study focuses on urban water metabolism. Thus, based on empirical fieldwork on the urban river Lea in East London, and supplemented by scientific literature and technical documents, I analyse three urban microbial ecologies that correspond to the urban realms ‘extended microbiomes’: those involved in slow sand filtration for the treatment of drinkable water, those involved in sewage treatment via the activated sludge process, and those emerging and evolving in disused urban canal infrastructure. These processes spatially manage microbial growth and modulate the distribution of different forms of microbial agency with important effects for the smooth functioning of urban water metabolism. Yet, considering the specific materiality and ecological agency of microbes points to a more-than-human contingency and indeterminacy at the centre of urban reproduction. Urban human-microbial co-evolution and is mired with uncertainties, ‘black boxes’, and unintended consequences.