Diffusive Gradient in Thin film (DGT) is a speciation technique now commonly used in the scientific literature to assess metallic contamination in water. It consists of several layers of chelating resin, hydrogel and protective membranes. The thickness of the layer of diffusion is a key parameter to evaluate the labile concentration of metals by DGT. As any solid immersed in water, there is a region around DGT where tangential velocity of water decreases, creating a diffusive boundary layer (DBL). At first, its thickness has been considered negligible compared to the gel layer in well-stirred solutions. But recent studies highlights that in water courses with low flow, DBL has to be considered because it can affect the evaluation of the labile concentration. All estimations of DBL are obtained with DGT equipped with classical diffusive gels. Briefly, it is the use of different thicknesses of gels, associated with the capacity of complexes of Cd, Pb, Zn to dissociate rapidly, that allows to calculate a value of DBL. The aim of the study was to determine DBL with another type of gel, which has more restricted pores, and to define a simple method to evaluate DBL in this configuration that we can apply to field data. The idea is that the complexes sampled are different, and the diffusion in the gel is slower, so we verified that the DBL established with restricted gels matched the model proposed by previous studies for diffusive gels. DBL has been assessed for Cd, Cr, Co, Cu, Mn, Ni, Pb, V, Zn at the same time, in controlled conditions in the laboratory, with varying stirring speed and water composition. Different values of DBL have been calculated. The robustness of the method has been then tested by applying it to mesocosm experiment, where DBL was also evaluated with DGT equipped with diffusive gels. The experimental system consisted in canals with different flow velocities receiving effluent from a waste water treatment plant. The best estimation of the physical DBL was given by Co in this case. Finally, the feasibility of calculating a DBL in real conditions of a large scale field deployment, with only two different thicknesses is discussed.