This dissertation aims to enrich understanding of issues affecting the care and protection of orphans and vulnerable children (OVC) in sub-Saharan Africa. The social risks and resources of OVC in selected contexts is examined using both quantitative and qualitative data and multiple datasets. Results pertaining to OVC's sexual risk, marginalization, available social support and reasons for limited support are presented. Chapter 1 investigates the onset of sexual behavior and sexual risk among orphans and non-orphans in South Africa. The sample for this analysis consists of 1694 Black African youth age 14-18. The analyses found both male and female orphans significantly more likely to have engaged in sex as compared to non-orphans (49% vs. 39%). Among sexually active youth, orphans reported younger age of sexual intercourse with 23% of orphans having had sex by age 13 or younger compared to 15% of non-orphans. Chapter 2 focuses on a sample of 692 youth-headed households (YHH) in Rwanda age 13-24 and describes their degree of marginalization and available social support. Most youth reported significant caring relationships: 73% reported access to trusted adult who offers them advice and guidance, and most indicated close peer relationships. However, many youth also perceived a lack of community support, with 86% feeling rejected by the community and 57% feeling the community would rather hurt them than help them. Chapter 3 expands upon the results presented in Chapter 2 to gain further insight into the marginalization and limited community support reported by YHH. The socio-cultural factors that influence the level of support Rwandan communities provide to OVC is explored through a triangulation of multiple methods and perspectives. The analyses cast light on the importance of three factors affecting community support and marginalization: stigma, NGO assistance and community discord. Overall, this dissertation enriches understanding of how OVC's vulnerability and social networks impact their care and protection. Data such as this should be used to inform efforts to support OVC. In hopes of achieving this aim, discussion throughout lends particular attention to the programmatic implications of these results
