Abstract In this study, a computational fluid dynamics (CFD) method was adopted to calculate axial dispersion coefficients of annular pulsed disc and doughnut columns (APDDCs). Passive tracer was uniformly injected by pulse input at the continuous phase inlet, and its concentration governing equation was solved in liquid–liquidtwo-phase flow fields. The residence time distributions (RTDs) were obtained using the surface monitoring technique. The adopted RTD–CFD method was verified by comparing the axial dispersion coefficient between simulation and experimental results in the literature. However, in pilot-scale APDDCs, the axial dispersion coefficients predicted by the CFD–RTD method were approximately three times larger than experimental results determined by the steady-state concentration profile method. This experimental method was demonstrated to be insensitive to the variation of the axial dispersion coefficient. The CFD–RTD method was more recommended to determine the axial dispersion coefficient. It was found that the axial dispersion coefficient increased with an increase in pulsation intensity, column diameter, and plate spacing, but was little affected by the throughput.