As part of a European LIFE ArtWET project, two on‐site buffer zones, an artificial wetland and a forest plot, are being evaluated for their capacity to mitigate pesticide pollution. As treatment efficiency is highly dependent on the systems' hydrology, the present work focuses on the watershed and both systems' hydrological functioning. The design strategy involved limited inlet flow rates to 70 L s−1: 99% of watershed outlet flow rates were lower than this limit. Approximately half of the flows of greatest concern passed through the artificial wetland, whereas the forest only received 2% of these flows. A tracer experiment was conducted under a low steady flow rate while little vegetation was present in the artificial wetland. A water dye tracer (sulforhodamine B, SB) and two molecules of contrasting properties, uranine (Ur, photodegrading) and isoproturon (mobile and only slightly sorptive, IPU) were injected. Dilution, sorption, and photodegradation were observed. The forest plot, which presented a high organic matter content, showed more sorption (IPU, SB) but lower photodecay (Ur) than did the artificial wetland. Total IPU losses in the forest buffer were high (79%). In the artificial wetland, 30% IPU losses were found, whereas a 66.5‐h mean retention time was determined and good hydraulic efficiency (0.55) was calculated. Few dead zones and short‐circuits were found, suggesting good hydrological functioning. Implementing buffer zones in subsurface pipe‐drained watersheds actively participates in the reduction of pesticide transfer to natural water bodies.