For many decades in India, natural resource management (NRM) programs were implemented by government bureaucracies in a centralized, top-down manner. The programs were unsustainable and suffered from resource use inefficiency and inequity. In the 1990s, under pressure from civil society organizations and multilateral agencies, the Government of India and many State Governments introduced policies that decentralized NRM programs and mandated active participation of users in the management of resources. When implementation responsibilities were transferred to resource user-groups many of the problems associated with centralization could be reduced significantly. However, despite their proven capacity of being better resource managers than government agencies, the user-groups encountered difficulties as self-governed people's organizations. Participation of users declined and problems of equity resurfaced in many user-groups. This dissertation describes the research that examines the causes of problems in the governance of user-groups in villages of Mehsana District in Gujarat. Using an eight-fold criteria of good governance, the study looks at eight Water Users Associations (WUAs) that took over irrigation management responsibilities from the Irrigation Department. This program of decentralization of irrigation is called Participatory Irrigation Management (PIM). The assessment of each WUA on each of the eight criteria reveals a close link between characteristics of good governance and the process of democratization. It can be seen that the WUAs that performed well on participation, equity, transparency, accountability, rule of law and consensus-orientation were less likely to face situations of dysfunction than the WUAs that performed poorly on these criteria. These criteria for good governance are also the core elements of democratic governance. At the same time, the case-studies reveal the tension between the democratization process that is attempted within the WUAs and the historical and cultural legacy of the feudal, autocratic and patriarchal society that rural India has been for many centuries. The thesis supports the argument, with empirical evidence, that the decentralization process can be sustainable only when user-groups institutionalize democratic processes and the early leaders behave in a democratic manner. It also suggests that the transition from an undemocratic institution to a democratic one can be enabled when external support agencies play an important catalytic role.