This thesis aims to explore the socio-spatial control mechanisms, termed as Informal land controls - used by informal power intermediaries termed informal structures to produce space to claim public resources in Karachi-Pakistan. Such informal structures include but are not limited to: political parties, bonds of tribe and kinship, religious authorities, neighbourhood committees, gangs and other criminals. The concept of informal land controls is integrally embedded within the processes of informality and production of space in Karachi and is used to examine political conflicts, social divisions and the spiral of violence evident in Karachi, especially from mid-2000 onwards. The study introduces the concept of informal land controls to address the gap in the literature, both on informality and production of space, which currently remains focussed on the state, market and the poor, and thus, neglects the role of power intermediaries. Such informal processes of production of space that are neither 'from above' i.e. state and market driven, nor 'from below', but 'from between'. The research is based on four case study areas in Karachi - North Nazimabad, Lyari, PIB Colony and Scheme 33 - which demonstrate how multiple and overlapping informal structures initiate and maintain informal land controls, and how these are linked with the failure of formal government and prevailing informality in planning decisions. A wide range of licit and illicit measures are used - including political domination; religious discourse; cultural practice; tribal identity; property control; barriers, flags and wall-chalking; harassment; extortion; violence, street control and homicides. The result is a complex network of socio-spatial divisions that form 'no-go' areas of varying degrees, which have exacerbated social divisions and violence in the city. The research argues that understanding such processes is essential to underpin interventions to reduce violence and extortion, and ensure more equitable access to urban resources.