Vast areas of (sub)arctic tundra are dominated by the ericoid dwarf shrub Empetrum hermaphroditum. Recent experimental and observational data have shown that Empetrum can be damaged heavily by recurrent extreme winter warming. In addition, summer warming leads to increased soil N availability in tundra ecosystems. In a 7-year experiment, I investigated the recovery of subarctic Empetrum-dominated tundra vegetation using a factorial combination of various degrees of aboveground Empetrum removal (simulating the damaging effects of extreme winter warming) and N addition (simulating one of the effects of summer warming). After 7 years no new species had established in the plots. The growth of planted Betula nana seedlings was stimulated by Empetrum removal and reduced by N addition. This Empetrum-dominated tundra ecosystem was resilient against severe disturbances. Only when Empetrum was 100% removed did it fail to recover, and only in combination with high N supply the subordinate species (notably Eriophorum vaginatum and Rubus chamaemorus, a graminoid and a forb) could benefit. In the 50% removal treatment Empetrum recovered in 7 years when no N was supplied and the cover of the subordinate species did not change. However, when N was added Empetrum recovered faster (in 4 years) and the subordinates decreased. When Empetrum was not removed and N was added, Empetrum even increased in abundance at the expense of the subordinate species. Thus, profound changes in tundra ecosystems can only be expected when Empetrum is very heavily damaged as a result of recurrent extreme winter warming and when soil N availability is increased as a result of summer warming. These changes in species composition upon extreme disturbance events may lead to a wide variety of ecosystem feedbacks and cascade processes as this tundra system is relatively species-poor, and can be hypothesized to have low functional redundancy.