Abstract Background Obstetric interventions performed during delivery do not reflect improvements in obstetric care. Several practices routinely performed during childbirth, without any scientific evidence or basis - such as Kristeller maneuver, routine episiotomy, and movement or feeding restriction - reflect a disrespectful assistance reality that, unfortunately, remains in place in Brazil. The aims of the current study are to assess the coexistence and prevalence of obstetric interventions in maternity hospitals in Belo Horizonte City, based on the Grade of Membership (GoM) method, as well as to investigate sociodemographic and obstetric factors associated with coexistence profiles generated by it. Methods Observational study, based on a cross-sectional design, carried out with data deriving from the study “Nascer em Belo Horizonte: Inquérito sobre o Parto e Nascimento” (Born in Belo Horizonte: Survey on Childbirth and Birth). The herein investigated interventions comprised practices that are clearly useful and should be encouraged; practices that are clearly harmful or ineffective and should be eliminated; and practices that are inappropriately used, in contrast to the ones recommended by the World Health Organization. The analyzed interventions comprised: providing food to parturient women, allowing them to have freedom to move, use of partogram, adopting non-pharmacological methods for pain relief, enema, perineal shaving, lying patients down for delivery, Kristeller maneuver, amniotomy, oxytocin infusion, analgesia and episiotomy. The current study has used GoM to identify the coexistence of the adopted obstetric interventions. Variables such as age, schooling, skin color, primigravida, place-of–delivery financing, number of prenatal consultations, gestational age at delivery, presence of obstetric nurse at delivery time, paid work and presence of companion during delivery were taken into consideration at the time to build patients’ profile. Results Results have highlighted two antagonistic obstetric profiles, namely: profile 1 comprised parturient women who were offered diet, freedom to move, use of partogram, using non-pharmacological methods for pain relief, giving birth in lying position, patients who were not subjected to Kristeller maneuver, episiotomy or amniotomy, women did not receive oxytocin infusion, and analgesia using. Profile 2, in its turn, comprised parturient women who were not offered diet, who were not allowed to have freedom to move, as well as who did not use the partograph or who were subjected to non-pharmacological methods for pain relief. They were subjected to enema, perineal shaving, Kristeller maneuver, amniotomy and oxytocin infusion. In addition, they underwent analgesia and episiotomy. This outcome emphasizes the persistence of an obstetric care model that is not based on scientific evidence. Based on the analysis of factors that influenced the coexistence of obstetric interventions, the presence of obstetric nurses in the healthcare practice has reduced the likelihood of parturient women to belong to profile 2. In addition, childbirth events that took place in public institutions have reduced the likelihood of parturient women to belong to profile 2. Conclusion(s) Based on the analysis of factors that influenced the coexistence of obstetric interventions, financing the hospital for childbirth has increased the likelihood of parturient women to belong to profile 2. However, the likelihood of parturient women to belong to profile 2 has decreased when hospitals had an active obstetric nurse at the delivery room. The current study has contributed to discussions about obstetric interventions, as well as to improve childbirth assistance models. In addition, it has emphasized the need of developing strategies focused on adherence to, and implementation of, assistance models based on scientific evidence.