Objectives We examined stillbirths in relation to disinfection by-product (DBP) exposures including chloroform, bromodichloromethane (BDCM), dibromochloromethane, bromoform, trichloroacetic acid, dichloroacetic acid (DCAA), monobromoacetic acid and summary DBP measures (trihalomethanes (THM4), haloacetic acids (HAA5), THMBr (brominated trihalomethanes) and DBP9 (sum of THM4 and HAA5)).

Methods We randomly selected 10 controls for each of the 2460 stillbirth cases with complete quarterly 1997–2004 THM4 and HAA5 town-level drinking water data. Adjusted (aORs) were calculated based on weight-averaged second-trimester DBP exposures.

Results We detected statistically significant associations for stillbirths and the upper DCAA quartiles (aOR range: 1.50–1.71). We also found positive associations for the upper four HAA5 quintiles and different stillbirth cause of death categories that were examined including unexplained stillbirth (aOR range: 1.24–1.72), compression of umbilical cord (aOR range: 1.08–1.94), prematurity (aOR range: 1.37–2.88), placental separation and haemorrhage (aOR range: 1.44–2.01) and asphyxia/hypoxia (aOR range: 1.52–1.97). Additionally, we found positive associations between stillbirths and chloroform exposure (aOR range: 1.29 – 1.36) and unexplained stillbirths and BDCM exposure (aOR range: 1.51 – 1.78). We saw no evidence of exposure–response relationships for any categorical DBP metrics.

Conclusions Consistent with some previous studies, we found associations between stillbirths and chloroform and unexplained stillbirth and BDCM exposures. These findings strengthen existing evidence of prenatal THM exposures increasing the risk of stillbirth. Additionally, we saw statistically significant associations between DCAA and stillbirth. Future research should examine cause-specific stillbirths in relation to narrower critical windows and additional DBP exposure metrics beyond trihalomethanes and haloacetic acids.