In order to reveal the combined pollution characteristics of malathion adsorption onto the surficial sediments with pesticides (dimethoate, metalaxyl, atrazine, malathion, prometryn) and heavy metals (Cu, Zn, Pb, Cd, Ni) present, a 2^10-5 fractional factorial design of resolution IV was employed to identify non-interactive effects and two-way interactions of the chosen factors on malathion adsorption. The missing values of malathion adsorption in the experiment were approximated using the algorithm of back propagation neural network. It was found that Zn (13.69%), atrazine (8.76%), and metalaxyl (5.99%) have significant positive effect (α=0.05) on the malathion adsorption; 44.92% of the total contribution to malathion adsorption came from the non-interactive effects, while 55.08% was from the two-way interactions (Cu×metalaxyl, Zn×atrazine, Cu×prometryn, Pb×Cu, Cu×atrazine, and Cu×Ni). Notably in the two-way interactions, Zn×atrazine (12.93%) and Cu×prometryn (5.97%) have a strong synergistic effect on malathion adsorption, while Pb×Cu (15.80%), Cu×atrazine (7.82%), Cu×metalaxyl (6.63%), and Cu×Ni (5.93%) present intense antagonisms to malathion adsorption. The whole philosophy could be used as a new methodology in accessing the degree of combined pollution and the effect of multiple pollutants on the selected pollutants.