Evidence from cold regions in North America has shown that the performance of stormwater ponds differs between winter and summer. The pond hydraulics change seasonally, and winters have lowered removal efficiency due to a combination of an ice cover, cold water and de-icing salts. This study examines the function of the Bckaslov stormwater pond under the more mild conditions of southern Sweden, where there are several snow and melt cycles per year.Event sampling in the summer of 1997 showed good removal efficiencies for nutrients, total suspended solids (TSS) and a selection of metals (Cd, Cu, Pb, Zn), but winter grab-tests taken in 1995–96 and 1997–98 suggest that the pond acts as a pollutant source under cold conditions. To better assess winter and spring pond performance, water at the inflow and outflow was sampled from January to April 2003. The low intensity of runoff delivery and slow inflow velocities meant that time- rather than flow-weighted sampling was used. Five consecutive events were sampled and analysed for TSS, chloride and the metals As, Cd, Cr, Cu, Hg, Ni, Pb and Zn. YSI probes were in place at both the inlet (pH, temperature) and outlet (pH, temperature, conductivity, dissolved oxygen) to determine the timing of pollution flows. In addition, profiles of the same quality indicators allowed snapshots of pond processes.De-icing salt has a major effect on pond hydraulics. Strong stratification occurred after each snowmelt-generated flow event and up to 80% of chloride could be retained by the pond. However, continuous conductivity measurements show that chloride is flushed between events. Ice changes retention times and causes oxygen depletion, but bed scour was not observed. Pond performance decreased during the winter and spring, albeit not as badly as the grab tests suggest. A seasonal comparison of the removal efficiencies showed that removal of Cd (75%) and Cu (49%) was about the same for summer and winter–spring, but removal of Pb, Zn and TSS dropped from 79%, 81% and 80% to 42%, 48% and 49% respectively. The removal efficiencies for the other metals sampled in 2003 were: As, 50%; Cr, 39%; Hg, 56%; Ni, 41%. Copyright © 2005 John Wiley & Sons, Ltd.