Bisphenol S (BPS) and bisphenol F (BPF) are increasingly used in manufacturing consumer products to replace the use of bisphenol A (BPA), but exposure data are limited, particularly among pregnant women. Here, we measured BPA, BPS, and BPF levels in urine samples, collected from 941 pregnant women over three trimesters. We examined the correlations, coexposure patterns, variability, and predictors of bisphenols using Spearman's correlation coefficient, percentile analysis, intraclass correlation coefficient, and linear mixed models, respectively. We assessed health risks using average concentrations of bisphenols over three trimesters. The three bisphenols were detected in more than 50% of samples, among which BPA was the predominant one. Cashiers, office workers, teachers, and salespersons had elevated urinary BPS concentrations, while healthcare workers had relatively higher BPA concentrations. About 15 participants had potential health risks induced by exposure to bisphenol mixtures. These findings indicate that exposure to multiple bisphenols at low levels is common over three trimesters. Multiple measurements of urinary BPA and BPS concentrations are needed for more accurate evaluation of the exposure levels during pregnancy, while urinary BPF concentrations during pregnancy are moderately reliable. Occupational exposure should be taken into consideration in future demographic studies.