A membrane of triolein-embedded cellulose acetate membrane (TECAM), as an integrative passive sampler, was applied to adsorb 28 polychlorinated biphenyls (PCBs) from pure water and sea water in order to probe into the influence of concentration fluctuation and water matrix. The results demonstrated the temporal variations of PCBs concentrations in TECAM followed the first-order kinetics model. The periodic refreshment of solution and matrix effects of sea water significantly prolonged the time that PCBs reached equilibrium stage. The refreshment facilitated the uptake mass in TECAM. On the contrary, the matrix effects of sea water and dissolved organic matter (DOM) declined the PCBs absorption to TECAM. The average logKP values of PCBs in pure water were about 1.2 log unit higher than those in sea water in the experiments that the solution was not refreshed, while the difference of average logKP values narrowed to 0.3 log unit if the PCBs solutions were periodically refreshed. The correlation between logKP and logKOW values fitted the quadratic curve well, which was similar to semi-permeable membrane device (SPMD). The appropriate sampling times (t94%) ranged from 98.8 to 819 h (mean 500 h) for pure water with refreshment, much longer than those in sea water with refreshment (80.1~410 h, mean 189 h). The t94% values in the solution with high DOM content increased significantly, up to 409 h. Furthermore, comparing the twoexperiments that the spiked pure water and sea watersolution were refreshedfrequently, the estimated sampling rates (Rs) in pure water (0.154~2.06 L/day with a mean value of 0.605 L/day) were slightlylower than those in sea water (0.292~3.84 L/day with a mean value of 1.69 L/day). However, the Rs values in sea water with DOM declined sharply to 0.042 L/day. Therefore, concentration fluctuation, matrix effect, and DOM contents of sea water evidently posed significant influence on dynamic parameters of TECAM absorption, which would be screened and probed detailed in future.