Correctly assessing the amount of blood loss is crucial in order to adequately treat postpartum haemorrhage (PPH) at an early stage and diminish any related symptoms and/or complications. The aim of our study is to analyse correctness in visually estimated blood loss during labour and to measure the differences between subjectively measured and weighted blood losses (ml). Cross-sectional study A Swedish maternity unit with 6000 annual births Midwives employed at a big maternity unit at a hospital in northern Stockholm, Sweden. Midwives assisting 192 vaginal births were asked to visually estimate the blood loss from the assisted delivery. Coasters and sanitary pads were weighed following the birth. We analysed if there were any differences between subjective measured blood loss (ml) and weighted blood loss. These two methods were also compared to quantify concordance between estimated blood volume and the actual volume. The number of overestimates of blood loss was 45.3 % (n=87) with an average of 72.9 ml; the number of underestimates was 49.4 % (n=95) with an average of 73.8 ml. Exact correct estimations of blood loss were done in 5.2 % of the cases (n=10). The largest overestimation of a postpartum bleeding was by 520 ml; the largest underestimation was by 745 ml. There was both underestimation and overestimation of blood loss. We found small but significant overestimates in PPH < 300 ml (16 ml). In PPH > 300 ml, there was a small but not significant underestimates (34 ml). Based upon our findings, we conclude that it is reasonable to start weighing blood loss when it exceeds 300 ml. •Both underestimating and overestimating blood loss was done.•Exact correct estimates of PPH were done in 5.2 % of the cases (n=10).•In 83 % of the cases (n=160), the absolute estimation error is less than 120 ml.