BackgroundLow rates of maternal healthcare service utilization, including facility delivery, may impede progress in the prevention of mother-to-child transmission of HIV (PMTCT) and in reducing maternal and infant mortality. The MoMent (Mother Mentor) study investigated the impact of structured peer support on early infant diagnosis presentation and postpartum maternal retention in PMTCT care in rural Nigeria. This paper describes baseline characteristics and correlates of facility delivery among MoMent study participants.MethodsHIV-positive pregnant women were recruited at 20 rural Primary Healthcare Centers matched by antenatal care clinic volume, client HIV prevalence, and PMTCT service staffing. Baseline and delivery data were collected by participant interviews and medical record abstraction. Multivariate logistic regression with generalized estimating equation analysis was used to evaluate for correlates of facility delivery including exposure to structured (closely supervised Mentor Mother, intervention) vs unstructured (routine, control) peer support.ResultsOf 497 women enrolled, 352 (71%) were between 21 and 30 years old, 319 (64%) were Christian, 245 (49%) had received secondary or higher education, 402 (81%) were multigravidae and 299 (60%) newly HIV-diagnosed. Delivery data was available for 445 (90%) participants, and 276 (62%) of these women delivered at a health facility. Facility delivery did not differ by type of peer support; however, it was positively associated with secondary or greater education (aOR 1.9, CI 1.1–3.2) and Christian affiliation (OR 1.4, CI 1.0–2.0) and negatively associated with primigravidity (OR 0.5; 0.3–0.9) and new HIV diagnosis (OR 0.6, CI 0.4–0.9).ConclusionsPrimary-level or lesser-educated HIV-infected pregnant women and those newly-diagnosed and primigravid should be prioritized for interventions to improve facility delivery rates and ultimately, healthy outcomes. Incremental gains in facility delivery from structured peer support appear limited, however the impact of duration of pre-delivery support needs further investigation. Religious influences on facility delivery and on general maternal healthcare service utilization need to be further explored.Trial RegistrationClinicalTrials.gov number NCT01936753, registered September 2013.Electronic supplementary materialThe online version of this article (doi:10.1186/s12884-017-1417-2) contains supplementary material, which is available to authorized users.