ABSTRACT OBJECTIVE To analyze if the presence of a companion favors the use of best practices in the delivery care in the South region of Brazil. METHODS This is a cross-sectional analysis of the longitudinal study Nascer no Brasil. We analyzed data from 2,070 women from the South region of Brazil who went into labor. The data were collected between February and August 2011, by interviews and medical records. We performed a bivariate and multivariate analysis, calculating the crude and adjusted prevalence ratios using Poisson regression with robust variance estimation. The level of significance adopted was 5%. RESULTS Most women had a companion during labor (51.7%), but few remained during delivery (39.4%) or cesarean section (34.8%). Less than half of the women had access to several recommended practices, while non-recommended practices continue to be performed. In the model adjusted for age, education level, source of payment for the delivery, parity, and score of the Brazilian Association of Market Research Institutes, the presence of a companion was statistically associated with a greater supply of liquids and food (aPR = 1.34), dietary prescription (aPR = 1.34), use of non-pharmacological methods for pain relief (aPR = 1.37), amniotomy (aPR = 1.10), epidural or spinal analgesia (aPR = 1.84), adoption of non-lithotomy position in the delivery (aPR = 1.77), stay in the same room during labor, delivery, and postpartum (aPR = 1.62), skin-to-skin contact in the delivery (aPR = 1.81) and cesarean section (PR = 2.43), as well as reduced use of the Kristeller maneuver (aPR = 0.67), trichotomy (aPR = 0.59), and enema (aPR = 0.49). CONCLUSIONS In the South region of Brazil, most women do not have access to the best practices in addition to undergoing several unnecessary interventions. The presence of a companion is associated with several beneficial practices and the reduction in some interventions, although other interventions are not impacted.