Due to the entrance of pollutants in different branches of the river network, it is essential to study contaminant transport at the river confluences. In the present study, it was attempted to investigate the conservative pollution transport at channel confluence by operating a series of experiments in the laboratory flume. In the designed laboratory model, two branches, with different widths of 45, 25 cm, were intersected and a channel confluence was created. Five entrance discharges and three initial contaminant concentrations, introduced using a linear feeder, were chosen as experimental variables. Conservative tracer of sodium chloride solution was used, and the electrical conductivities were measured at eight locations of the main channel and upstream branches with 2 seconds interval. Junction zone was assumed as a control volume, and by applying mass equilibrium to it, a new mathematical model was extracted. It was observed that there is concentration fluctuation in the falling limbs of the experimental breakthrough curves of the junction zone; however, it was diminished by downstream motion. Moreover, the observed pollution graphs had double peak points which changed to a single point with an increase of distance from the confluence position. Operation of the presented model was investigated by variation of its parameters. It was found that the contaminant residence time parameters of the confluence zone have the most significant influence in the simulation of the analytical model. Additionally, it was observed that the values of Gaussian distribution of the upstream branches could displace the position of pulses of resultant breakthrough curves or can overlap them. Moreover, the model performance was examined using statistical goodness of fit parameters like Nash-Sutcliffe, R2, and mean absolute error (MAE). Their values were calculated as 0.88, 0.91, 66.88 (ppm), respectively.