Master of Public Health - MPH / Two quantitative studies were carried out in randomly-selected facilities within all nine provinces of South Africa. First, a situational assessment of these randomly selected facilities was undertaken using key informant (health care personnel) interviews and record reviews to ascertain guidelines and procedures for early identification of HIV-exposed infants (HEI), the coverage of early infant diagnosis services, the human resource capacity of the health system, and existing linkage and referral system for antenatal and postnatal PMTCT services. This was followed by the South African national PMTCT survey (SAPMTCTE) which involved a collection of infant blood samples and maternal interview data from mother-infant pairs (infants age 4-8weeks) attending six weeks immunisation service points in the selected facilities. Interviews were conducted with mothers to assess antenatal and peripartum PMTCT services received and maternal intention to request for infant HIV testing at six weeks immunisation visits. Data on gestational age at birth, infant birth weight and HIV status was extracted from the road-to-health-card (RtHC). The HIV status of mothers was determined from maternal report or enzyme immunoassay (EIA) test conducted on infants dried blood spots (DBS). A weighted analysis (weighted for sample size realisation and population live births) was performed to assess uptake of services along the PMTCT cascade. Mothers who either self-reported an HIV-positive status or had an EIA positive infant were classified as HIV-positive mothers. Perinatal ARV regimen coverage was calculated from the total number of HIV-positive mothers who received maternal azidothymidine (AZT) or HAART for any duration during pregnancy plus infant nevirapine (NVP)/AZT received at birth. Descriptive methods were used to analyse national availability of EID services and approaches for identifying HEI at the six weeks immunisation visit. Logistic regression assessed key factors influencing maternal intention to receive EID. Logistic regression was also used to explore individual, health facility and provincial level factors that explain variability in mother-to-child-transmission rates.