Understanding the space-time structure and characteristics of drought is crucial for drought risk mitigation and forecasting efforts. In this paper, the drought events are identified by an improved 3-dimensional clustering algorithm. The 3-month Standardized Precipitation Evapotranspiration Index (SPEI3) is calculated based on the Climatic Research Unit (CRU) dataset from January 1966 to December 2015 to construct the longitude-latitude-time space. The characteristics of these identified events are described by a series of indicators such as duration, severity, intensity, affected area, centroids and track path after which the spatial pattern and temporal evolution are studied. In addition, the Vegetation Health Index (VHI), Vegetation Condition Index (VCI) and Temperature Condition Index (TCI) are also included to investigate the possible impact of drought on the vegetation health. Results show that a total of 60 persisting drought events for at least 3 months have been identified in Central Asia. The most severe drought event occurred between May 1974 and January 1976 with a duration of 21 months and an affected area of 5.77 million km². Most events have a duration between 3 and 5 months. In addition, the droughts in Central Asia tend to have an east-west trajectory. Seasonally, the study area experienced more drought months in spring and summer. During the recent 35 years, there has been a slowly increasing thermal stress and a significantly decreasing soil moisture stress that causes an overall decreasing drought stress on vegetation health.