A drought is an extreme moisture deficit event caused by meteorological factors that destroys the structure and function of ecosystems. The recovery time (RT) is a critical metric that describes the responses of ecosystems to drought. However, the factors influencing ecosystem RTs are still unclear at the seasonal scale, especially the influence of the climatic state and the biological processes behind it during the recovery period. We selected a severe drought that occurred in Northeast China (NEC) from October 2018 to April 2019 (Win18-Spr19) and observed the vegetation indicator dynamics during and after the drought. Hierarchical partitioning and partial correlation analysis were used to quantify the contributions of the influencing factors to the RT. Abnormally high temperatures, and low precipitation and snowfall led to this drought event, which caused greenness loss in more than half of the ecosystems. Nearly half of the ecosystems recovered within 4 months, while regions with RTs longer than 9 months were concentrated in high-latitude regions, accounting for approximately 6% of the analysed ecosystems. The RTs were significantly negatively correlated with temperature anomalies in the high-latitude regions of NEC. This result indicated that ecosystems that experienced greater temperature reductions tended to have longer RTs, reflecting the negative impacts of low growing season temperatures on ecosystem recovery in the high-latitude regions of NEC. The abnormal climate pattern in which a warmer winter is followed by a colder summer threatens ecosystems: the vegetation loss caused by the warm winter drought event and slowing growth due to low summertime temperature. This climate pattern places ecosystems under unfavorable environmental conditions over long durations, including both the drought period and the long subsequent recovery period.