I Kamisan Atan and co-authors are to be congratulated for their rigorously designed and performed randomised controlled trial (RCT) of an interesting and clinically important question: can progressive vaginal distension, by use of an inflated vaginal balloon device in late pregnancy, decrease the occurrence of birth trauma? Unfortunately, the answer appears to be ‘no’. The pathophysiology of pelvic organ prolapse is becoming better understood, particularly with adva-nces in multidimensional magnetic resonance imaging (MRI) and ultrasound imaging (Singh et al. Am J Obstet Gynecol 2003;188:910–5) that detect subclinical pelvic changes related to vaginal birth, which has long been recognised as an important risk factor for prolapse. To allow passage of the evolutionarily large fetal head through a woman's pelvis, the pelvic muscles must undergo dramatic deformation, so it is not surprising that pelvic muscle damage, either avulsion or over-distension, occurs in as many as one-third of women having their first vaginal birth (Dietz et al. BJOG 2008;115:979–84). Similarly, results of this RCT found pelvic muscle damage in 27% of women, with levator avulsion in 13.1% and over-distension (termed microtrauma in this trial) in 13.8%. In contrast to the results of a small observational study that reported a significantly higher chance of delivery with an intact perineum in women who used the Epi-No device, compared with standard practice (Kovacs et al. Aust NZ J Obstet Gynaecol 2004;44:347–48), perineal outcomes in this trial did not differ between groups. Overall, approximately three-quarters of women achieved vaginal delivery, with half delivering over an intact perineum and about one-quarter undergoing episiotomy. The trial results confirmed the under-diagnosis of obstetric anal sphincter damage, which was clinically recognised in only 5–7% of women, equally in both groups, versus 17.6% found by ultrasound. Overall, the RCT results were reassuring regarding safety: the use of the birth trainer did not seem to cause pelvic damage, with the possible exception of an increased risk of ultrasound-diagnosed external anal sphincter defect, which was marginally statistically significant in the intention-to-treat analysis (21% in the intervention group versus 14% in the control group; risk ratio, RR, 1.44; P = 0.07), and statistically significant in the treatment-received analysis (22 versus 12%, respectively; RR 1.76; P = 0.007). Is it possible, then, to alter the biomechanical properties of the pelvic muscles to protect them from damage during vaginal birth? This trial was not designed to answer that specific question; however, I believe the authors appropriately chose to study the clinical impact of their intervention. If any impact had been observed, it would then be possible to establish the mechanism by which the impact occurred. Negative results may seem disappointing, but they are an important contribution to the growing body of knowledge about birth trauma and later consequences for pelvic floor function. Birth-related pelvic damage cannot currently be predicted and, at least for levator avulsion, once such damage has occurred, it appears to be irreversible. Thus, ideally, the results of further research will inform changes in prenatal, obstetric, and postnatal practices to improve pelvic health, and its related quality-of-life impact, for women. None declared. Completed disclosure of interests form available to view online as supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.