Due to its workings as a global financial nexus, activists critical of capitalism have used the City of London to stage a number of large-scale direct actions since the 1983-84 Stop the City protests. By examining protest at this renowned site of intensified observation, I argue, we can learn a great deal about what surveillance processes do in practice. To develop its argument, this thesis offers a detailed examination of visual surveillance and counter surveillance practice over four protests: the J18 (1999); the G20 Meltdown (2009); Climate Camp in the City (2009); and Occupy LSX (2011). Based on empirical, qualitative research through archival work, interviews, and video documentation stored at the MayDay Rooms, this thesis demonstrates how City and Met police used visual surveillance to disrupt, re-frame and further criminalise dissent. Over the course of these four protests the police learnt new ways to suppress what they termed ‘extreme’ protest. Conversely, activists developed choreographed, embodied movements and alternative technologies to counter new public order procedures and police surveillance. Politically driven artists, performers and technologists were at the vanguard of these new protest formations, early internet livestreaming and pioneering technical innovations that challenged existing surveillant structures. Yet, as this thesis articulates, over the course of these protests many activists’ inventions were slowly subsumed into proprietary online frameworks, which embed surveillance by default. This thesis uses insights from Marx and Marxist inspired theorists to describe how this method of surveillance and subsumption took place. While police formations informed by this history are increasingly being taken up nationally and internationally, it is vital to understand how state security forces and corporate observers have dealt with ‘extreme’ protests in the City.