Over the last years, polymeric membranes have been used as passive samplers to assess hydrophobic organic contaminant concentrations in aquatic environments. The amount of contaminant accumulated in situ during long exposure periods enables the determination of a so called time-weighted average dissolved concentration in water, which is supposed to assimilate the contamination variations over time (Huckins et al., 1993). The objectives of this study were (i) to show if polymeric membranes are reliable to quantify micropollutants in the receiving environment by comparing the actual exposure concentration to the mean concentration estimated by the sampler; (ii) to understand how polymeric membranes assimilate a concentration that varies over time. After being spiked with PRCs, two types of polymeric membranes (LDPE and PDMS) were deployed during 21 days in channels bypassing the French river Gave in Lacq. The flow rate was controlled in the channels and three scenarios of pollution were tested. The first scenario reproduced chronic pollution by continuous injection of PAHs (anthracene, chrysene and benzo[a]pyrene). Two levels were tested: Environmental Quality Standard (EQS) level and 1/3 of the EQS level. The second scenario reproduced an accidental pollution by an injection of PAHs in the river during the first 3 days of the 21-day exposure. The third scenario was a discontinuous pollution with injection of PAHs during 3 periods of 3 days separated by 4 days. For the two latest scenarios, the same quantity of PAHs was released than during the first scenario. A duplicate of each sampler were retrieved at various time intervals to follow the PAH accumulation and the PRC elimination. During the experiment, grab samples of water were regularly analysed. PRC elimination and PAH accumulation were simultaneously observed for all the compounds allowing the determination of the time-weighted average dissolved concentrations in water for each duration of exposure. For the first scenario with continuous injections of PAHs, the concentrations calculated from passive samplers were the same for all exposure durations and consistent with the quantity of PAHs released in the channel. These results validated the laboratory calibration of kinetic parameters (presented in another IPSW abstract: Lorgeoux et al. 2013) and the use of PRC to correct in situ environmental exposure conditions. For the second and third scenarios with discontinuous injections, the accumulated quantities of PAHs followed the means of exposure concentrations. For the heaviest PAHs (chrysene and benzo[a]pyrene) the calculated concentrations were consistent with the average concentrations of grab samples. This demonstrates that polymeric membranes really incorporate changes in concentration over time for these molecules. For anthracene, the results also were conclusive for the scenario with 3 injections of 3 days. But for a single injection during the first 3 days of the 21-day period, most anthracene were desorbed leading to a measured concentration underestimated compared to the theoretical value. For PAHs of low molecular weight, the exchange kinetics seems too fast to assimilate a pollution peak in the time-weighted concentration.