Knowledge of the spatial and temporal distribution of pesticide concentrations is essential for pesticide registration. In field experiments performed during the registration procedure, suctions cups are widely used to monitor the evolution of pesticide leaching over time and to calculate the mean concentration for a given drainage period. Until now, there has been no detailed information regarding whether soil water sampling by suction cups is suitable for accurately predicting the field-scale leaching of pesticides in heterogeneous soils. To address this, we performed a numerical study to evaluate the uncertainties associated with suction cup sampling in heterogeneous soils under atmospheric boundary conditions using two pesticides with contrasting sorption and degradation properties. To address the two main operation modes of suction cup sampling, continuous and weekly sampling were analyzed. First, the variability of pesticide breakthrough concentrations was analyzed for a single pesticide application in a physically heterogeneous soil profile. Second, a 10-yr time series was used to analyze repeated pesticide applications, and estimate variability in the leached mass fraction (LMF) and mean concentrations. In both cases, pesticide breakthrough was compared to the breakthrough of an inert tracer. The results indicate that for a single pesticide application, pore water velocities varied only slightly for all operation sampling modes and compounds. On the other hand, the total extracted mass and LMF varied greatly, suggesting that the acceleration of the breakthrough sampled by suction cups plays an important role in terms of LMF and the mean concentration, especially for degradable substances.