It is now well-accepted that street-level bureaucrats (SLBs) play a key role as the face of government for the public and that their implementation actions exert immediate, major implications for citizens-clients. An extensive scholarly attention has been devoted to the ways through which SLBs exercise their discretion during direct delivery interactions . However, citizens are traditionally referred to as subjective to the actions of SLBs and referred to as the powerless side of the interaction. To allow a broader perspective on the role citizens play in their encounters with government, this dissertation focuses on citizen agency during street-level implementation and public service delivery. The main research question this thesis tries to answer is: how citizen agency during street-level implementation and public service delivery can be conceptualized, how does policy structure enable it, and what are its effects? To answer this question, I use three papers: one theoretical and two empirical. In the first paper I conduct a systematic literature review to know how literature has studied and defined citizen agency. In the second and third paper I explore the role of policy structure as an enabler of citizen agency, and particularly, the role of interactional structure. I use the empirical case study of Prospera, a conditional cash transfer in Mexico. The second paper contributes to answering my research question by focusing specifically on how the policy structure helps to develop citizen agency. This study fills a gap in the literature because it explains how policy structure contributes to citizen agency beyond individual factors like traditionally has been the case. In the third paper, I focus on repeated interactions between citizens and street-level bureaucrats as a source of citizen agency. Whit this study, I contribute to the literature in two ways: first, by providing a distinction between one-shot and repeated interactions, which the literature has avoided, and stating the possible consequences not only for the interaction but for street-level work in general. Second, by exploring how repeated interactions have implications for the way citizens behave during policy implementation and public service delivery.