The timing and duration of snow cover critically affect surface albedo, surface energy budgets, and hydrological processes. Previous studies using in-situ or satellite remote sensing data have mostly been site-specific (Siberia and the Tibetan Plateau), and remote sensing and/or modeling data include large uncertainties. Here, we used 1103 stations with long-term (1966-2012) ground-based snow measurements to investigate spatial and temporal variability in snow cover timing and duration and factors impacting those changes across the Eurasian continent. We found the earliest annual onset and latest disappearance of snow cover occurred along the Arctic coast, where the long-term (1971-2000) mean annual snow cover duration (SCD) was more than nine months which was the longest in this study. The shortest SCD, =10 days, was found in southern China. The first and last dates of snow cover (FD and LD, respectively), SCD, and the ratio of SCD to snow season length (RDL) were generally latitude dependent over the Eurasian Continent, while were elevation dependent on the Tibetan Plateau. During the period from 1966 through 2012, FD delayed and LD advanced by ~1 day/decade, and RDL increased by about 0.01/decade. The LD, SCD, and RDL anomalies (relative to the period 1971-2000) were also significantly correlated with latitude. Advances in LD and positive RDL were more significant in low-latitude regions, decreases in SCD were more significant in high-latitude regions. Changes in SCD were related to air temperature and snowfall in autumn and warming in spring. SCD specifically increased in the northern Xinjiang and northeastern China due to increased snowfall. The significant reduction in SCD in southwestern Russia, the Tibetan Plateau and along the Yangtze River was mainly affected by climate warming.