The presence of pesticides in streams in winter, five to six years following bans on their municipal use suggests that complicated transport behaviour, such as subsurface retention and/or accumulation of pesticides and its release during storms, could be important for understanding recovery time frames following bans or legislation that aim to reduce chemical inputs. We investigated late fall and winter dynamics of four herbicides in paired urban and rural watersheds in Toronto, Canada during rainfall and snowmelt. The range of average concentrations and loads of the sum of atrazine, metolachlor, 2,4-D and mecoprop overlapped in the two types of watersheds, with slightly higher average concentrations in the rural watershed. Relatively consistent herbicide concentration-discharge patterns (i.e. dilution) were observed in the urban sub-watersheds during rainfall, while concentration-discharge patterns were much more variable in the rural watershed. This suggests relatively uniform transport pathways across the urban sub-watersheds, compared to the rural watershed. Concentration-discharge patterns of the neutral herbicides atrazine and metolachlor were similar in both watersheds during snowmelt, though varying discharge patterns resulted in divergent timings of peak concentrations. In contrast, the acidic pesticides 2,4-D and mecoprop, which are primarily associated with urban uses, showed much more variable behavior across both watersheds and merit further investigation. Overall, this work highlights the need to consider pesticide dynamics throughout the year in order to more thoroughly assess the long-term efficacy of legislation governing their use.