Polar regions are predicted to undergo large increases in winter temperature and an increased frequency of freeze–thaw cycles, which can cause ice layers in the snow pack and ice encasement of vegetation. Early or late winter timing of ice encasement could, however, modify the extent of damage caused to plants. To determine impacts of the date of ice encasement, a novel field experiment was established in sub-arctic Sweden, with icing events simulated in January and March 2008 and 2009. In the subsequent summers, reproduction, phenology, growth and mortality, as well as physiological indicators of leaf damage were measured in the three dominant dwarf shrubs: Vaccinium uliginosum, Vaccinium vitis-idaea and Empetrum nigrum. It was hypothesised that January icing would be more damaging compared to March icing due to the longer duration of ice encasement. Following 2 years of icing, E. nigrum berry production was 83% lower in January-iced plots compared to controls, and V. vitis-idaea electrolyte leakage was increased by 69%. Conversely, electrolyte leakage of E. nigrum was 25% lower and leaf emergence of V. vitis-idaea commenced 11 days earlier in March-iced plots compared to control plots in 2009. There was no effect of icing on any of the other parameters measured, indicating that overall these study species have moderate to high tolerance to ice encasement. Even much longer exposure under the January icing treatment does not clearly increase damage.