Background: A large proportion of the rise in prematurity worldwide is owing to late preterm births, which may be due to the expansion of obstetric interventions, especially pre-labour caesarean section. Late preterm births pose similar risks to overall prematurity, making this trend a concern. In this study, we describe factors associated with provider-initiated late preterm birth and verify differences in provider-initiated late preterm birth rates between public and private health services according to obstetric risk. Methods: This is a sub-analysis of a national population-based survey of postpartum women entitled “Birth in Brazil”, performed between 2011 and 2012. We included 23,472 singleton live births. We performed non-conditional multiple logistic regressions assessing associated factors and analysing differences between public and private health services. Results: Provider-initiated births accounted for 38% of late preterm births; 32% in public health services and 61% in private health services. They were associated with previous preterm birth (s) and maternal pathologies for women receiving both public and private services and with maternal age 35 years for women receiving public services. Women receiving private health services had higher rates of provider-initiated late preterm birth (rate of 4.8%) when compared to the ones receiving public services (rate of 2.4%), regardless of obstetric risk–adjusted OR of 2.3 (CI 1.5–3.6) for women of low obstetric risk and adusted OR of 1.6 (CI 1.1–2.3) for women of high obstetric risk. The high rates of provider-initiated late preterm birth suggests a considerable potential for reduction, as such prematurity can be avoided, especially in women of low obstetric risk. To promote healthy births, we advise introducing policies with incentives for the adoption of new models of birth care.