Abstract Health services and especially community-based health services in Africa are characterized by inefficiency, limited capacity, and poorly-trained personnel. The quality of care in these systems is dependent most heavily on personnel: training, monitoring, and ongoing quality of the implementation. South Africa is deploying 65,000 community health workers (CHW), however, there are almost no systems currently, nor are the systems yet planned for how to train, monitor, and ensure quality care is delivered by these CHW. In a pilot over the last six months in the Eastern Cape of South Africa, we found 25% of CHW not showing up for work at any time, when supervision was out-sourced for the first time. Such performance and failures to monitor performance, with consequences of censuring non-attendance, are often typical of CHW service delivery in low and middle income countries (LMIC). Building on the results of a successful effectiveness study in the townships of Cape Town, South Africa, we aim to show that training, monitoring, and data-informed supervision will result in CHW having significant benefits in maternal and child health (MCH) over the first two years of life, particularly mothers living with HIV, depressed mothers, and teenage mothers. With support from the provincial government, Philani's innovative training and monitoring strategies aim to demonstrate that we can decrease the gap between what we know (based on science) and what we do to improve health in rural communities that currently lack accountability and efficiency. UCLA, Stellenbosch University, Zithulele Hospital, and the Philani Maternal, Child Health and Nutrition Trust will collaborate on this second effectiveness RCT of home visiting delivered by government-funded CHW. UCLA will randomize 24 matched areas, each with government-funded CHW to either: 1) home visiting trained, supervised, and routinely monitored (i.e., the Philani Intervention Model [PIM], n=420 women, 12 areas); or 2) a standard care (SC) control home visiting by CHW who receive current standards of training, monitoring, and supervision (n=420 women, 12 areas). Within the sample, 25% will be adolescent mothers; 30% will be MLH; 25% will drink alcohol in pregnancy; and 22% will be depressed. Independent assessments will be conducted by Stellenbosch University during pregnancy, within two weeks of post-birth, 6, 15, and 24 months later. We will construct an overall analytic test which examines multiple outcomes simultaneously with an omnibus test. We will monitor MLH's HIV testing, linkage to care, treatment adherence and retention; maternal health indicators (weight, diabetes, mental health status, partnerships, alcohol use, and partner violence); and, the child's physical growth, cognitive, affective, and behavioral adjustment. We expect PIM to significantly improve outcomes overall and for the high risk subgroups (MLH, depressed, adolescents, alcohol-using). Analyses of the real-time process measures ...