Abstract Background Intimate partner violence (IPV) against women is a serious health problem that affects pregnancy more frequently than other obstetric complications usually evaluated in antenatal visits. We aimed to estimate the accuracy of the Women Abuse Screening Tool-Short (WAST-Short) and the Abuse Assessment Screen (AAS) for the detection of IPV during and before pregnancy. Methods Consecutive eligible mothers in 21 public primary health antenatal care centres in Andalusia (Spain) who received antenatal care and gave birth during January 2017–March 2019, had IPV data gathered by trained midwives in the first and third pregnancy trimesters. The index tests were WAST-Short (score range 0–2; cut-off 2) and AAS (score range 0–1; cut-off 1). The reference standard was World Health Organization (WHO) IPV questionnaire. Area under receiver operating characteristics curve (AUC), sensitivity and specificity with 95% confidence intervals (CI) were estimated for test performance to capture IPV during and before pregnancy, and compared using paired samples analysis. Results According to the reference standard, 9.5% (47/495) and 19.4% (111/571) women suffered IPV during and before pregnancy, respectively. For capturing IPV during pregnancy in the third trimester, the WAST-Short (AUC 0.73, 95% CI 0.63, 0.81), performed better than AAS (AUC 0.57, 95% CI 0.47, 0.66, P = 0.0001). For capturing IPV before pregnancy in the first trimester, there was no significant difference between the WAST-Short (AUC 0.69, 95% CI 0.62, 0.74) and the AAS (AUC 0.69, 95% CI 0.62, 0.74, P = 0.99). Conclusions The WAST-Short could be useful to screen IPV during pregnancy in antenatal visits.