Various studies have dealt with the effects of physico-chemical parameters (temperature, pH, organic matter content, salinity) or physical parameters (flow velocity and temperature) under laboratory conditions for sampling of hydrophilic (pharmaceuticals and pesticides) chemicals. These studies have demonstrated i) variable influence of these parameters on sampling rates ii) some consecutive bias on the calculated time weighted average concentrations (TWAC). However, all of these factors were studied individually, and possible interactions were not addressed. The first part of our work was dedicated to the calibration of POCIS with various temperatures (10 and 25°C) and flow velocities (from 0.7-1 cm.s-1 to 44-45 cm.s-1). Principal effects, and possible interactions between the two parameters, on sampling rates (Rs) were investigated with a full factorial design of experiment. The results showed a predominant impact of the flow velocity on Rs for 10 moderately polar pesticides (log Kow=2.18-4.14). Conversely, the temperature seemed to have only secondary and low influence on Rs variations. The second part consisted in a compilation of Rs data for the 10 analytes calibrated during this study, and also reported in the literature. Knowing the exchanging surface area A for POCIS, the global resistance to the mass transfer coefficients 1/ko were determined from these Rs (with 1/ko=A/Rs), and then plotted against the experimental bulk flow velocity values reported. The fitting of a non-linear model based on the dimensionless Sherwood number provided useful estimates of regression coefficients, for a possible prediction of Rs under various flowing conditions. For instance, the results obtained in the case of atrazine (1/ko ranging from 0.054±0.003 and 0.132±0.027 d.cm-1, depending on flow velocity) were in good agreement with the 1/ko values estimated by Booij et al. 2017. Some correlations were also observed between the flow velocity and the desorption rate of DIA-d5 (from 0.023 to 0.154 d-1). Such an observation could be a promising way to take into account the DBL thickness contribution, according to the hydrodynamic conditions, and then a likely improvement of the accuracy for in situ TWAC estimates.