Poverty is a structural driver of risky sexual behaviors. While cash transfers can mitigate some of this risk, complementary interventions have been posited as a way to further reduce multidimensional vulnerability. We examine the impacts of a multicomponent intervention targeted to Tanzanian adolescents on their sexual behaviors and reproductive health. The intervention comprised livelihood and life skills training, mentoring, and health facilities’ strengthening. Data come from a cluster randomized controlled trial, where one study arm received the intervention and the other was randomized to control, but both arms participated in a government cash transfer program. Among 1,933 adolescents interviewed over three rounds, we found increases in contraceptive and HIV knowledge. The program also increased health seeking and HIV testing among boys, but slightly reduced age at sexual debut among girls. There were no impacts on contraceptive use, number of sexual partners, or pregnancy. Findings support the value of an adolescent intervention, and the fact that it was delivered within a social protection platform suggests a potential for scalability. Additional efforts are required to delay sexual debut and reduce the number of sexual partners and pregnancy, possibly through addressing supply‐side barriers and social norms, or through additional linkages to economic opportunities.

Introduction:Transactional sex or material exchange for sex is associated with HIV infection among adolescent girls and young women in sub-Saharan Africa. The motivations for engaging in transactional sex vary from the fulfilment of basic needs, to enhancing social status or for romantic reasons with the expectation that men should provide. Transactional sex is also associated with HIV risk behaviours, such as multiple sexual partners and other determinants of HIV risk, including partner violence and abuse, alcohol consumption and inconsistent condom use. Methods: We use data from a mixed-method, cluster randomised controlled trial of the Ujana Salama cash "plus" intervention in rural Tanzania. The data are from the first and third rounds of data collection (2017-2019). The impact evaluation consisted of a parallel mixed-methods design where the quantitative and qualitative data collection occurred simultaneously, and integration of the findings was done during the discussion. We first examine contextual factors associated with transactional sex using multivariable logistic regression models and then estimate whether the "plus" intervention reduced transactional sex among adolescent girls and young women using analysis of covariance. We used thematic content analysis for analysing qualitative transcripts. Results:The prevalence of transactional sex among unmarried adolescent girls and young women at round 3 was 26%. Findings show that increasing age is a risk factor for transactional sex (OR = 1.80; 95% CI: [1.50, 2.17]), staying in school was negatively associated with engagement in transactional sex (OR = 0.24; 95% CI: [0.14, 0.40]). The cash plus intervention showed no impacts on reducing transactional sex (β = 0.003, p = 0.905). Conclusions: The mechanisms of impact for a cash plus intervention on transactional sex are complex; economic insecurity is an important driver of transactional sex and HIV infection, but psychosocial factors and gendered social norms need consideration in intervention development. Our findings suggest that combination prevention interventions to address the structural drivers of HIV infection should focus on efforts to increase school enrolment and completion.