Present study aimed to generate multiple regression models to estimate the formation of trihalomethanes (THMs), haloacetonitriles (HANs) and haloacetic acids (HAAs) during chloramination of source water obtained from Yangtze River Delta Region, China. The results showed that the regression models for trichloromethane (TCM), dichloroacetonitrile (DCAN), dichloroacetic acid (DCAA), dihaloacetic acids (DHAAs), 5 HAAs species regulated by U.S. EPA (HAA5) and total haloacetic acids (HAA9) have good evaluation ability (prediction accuracy reached 81–94%), while the models for total haloacetonitriles (HAN4), trichloroacetic acid (TCAA), trihaloacetic acids (THAAs) and total trihalomethanes (THM4), they appeared relative low prediction accuracy (58–72%). For THMs, dissolved organic nitrogen (DON) was their key organic precursor, yet for HAN, DHAAs and THAAs, UVA254 played the dominant role. The other key factors influencing DBP formation included the bromide (THM4, DHAAs and HAA9), reaction time (DCAN, HAN4), chloramine dose (TCM, DCAA, TCAA, HAA5 and THAAs). These results provided important information for water works to optimize the water treatment process to control DBPs, and give an evaluating method for DBPs levels when estimating the health risks related with DBP exposure during chloramination. Display omitted •Regression models can well describe the DBPs formation during chloramination.•For THMs, dissolved organic nitrogen (DON) was their key organic precursor.•For HAN, DHAA and THAA, UVA254 played the dominant role.•Control of DON, UV254, Br- and NH2Cl dose may be effective to control DBPs.