Blood trihalomethanes (THMs) and urinary haloacetic acids (HAAs) are the leading candidate biomarkers for disinfection byproduct (DBP) exposure. However, no studies have assessed the exposure profiles, temporal variability, and potential predictors of these biomarkers during pregnancy. Here we collected blood (n = 4304) and urine samples (n = 4165) from 1760 Chinese pregnant women during early, mid-, and late pregnancy, which were separately analyzed for 4 THMs and 2 HAAs. We calculated the intraclass correlation coefficients (ICCs) to assess the variability of these biomarkers and estimated their correlations with sociodemographic, water-use behavioral, dietary and sample collection factors using mixed models. The median concentrations of TCM, BDCM, Br-THMs [sum of BDCM, dibromochloromethane (DBCM), bromoform (TBM)], total THMs (TTHMs, sum of TCM and Br-THMs), DCAA and TCAA in the water distribution system were 4.2 μg/L, 1.7 μg/L, 2.9 μg/L, 7.1 μg/L, 3.4 μg/L and 8.2 μg/L, respectively. Chloroform (TCM), bromodichloromethane (BDCM), dichloroacetic acid (DCAA) and trichloroacetic acid (TCAA) were detected in > 75% of the biospecimens. Repeated measurements of blood TCM, BDCM, Br-THMs and TTHMs and urinary DCAA and TCAA uniformly exhibited high variability (ICCs = 0.01–0.13); the use of a single measurement to classify gestational average exposure resulted in a high degree of exposure misclassification. The sampling season was a strong predictor of all analyzed DBPs. Additionally, we detected a positive association of blood TCM and BDCM with household income, urinary DCAA with age, and urinary TCAA with tap water usage, education level and amount of tap water consumed. Inverse associations were found between blood BDCM and vegetable consumption, and between blood Br-THM and TTHM and time interval since the last bathing/showering. Afternoon samples had lower DCAA concentrations than did early morning samples. Our results indicate that blood THM and urinary HAA concentrations vary greatly over the course of pregnancy and are affected by sampling season, time of day of blood/urine collection, sociodemographic factors, recent water-use activities and dietary intake.