In this paper, correlations between vegetation dynamics (represented by the normalized difference vegetation index (NDVI)) and hydro-climatological factors were systematically studied in Lake Baiyangdian during the period from April 1998 to July 2008. Six hydro-climatological variables including lake volume, water level, air temperature, precipitation, evaporation, and sunshine duration were ... [Show full abstract] used, as well as extracted NDVI series data representing vegetation dynamics. Mann-Kendall tests were used to detect trends in NDVI and hydro-climatological variation, and a Bayesian information criterion method was used to detect their abrupt changes. A redundancy analysis (RDA) was used to determine the major hydro-climatological factors contributing to NDVI variation at monthly, seasonal, and yearly scales. The results were as follows: (1) the trend analysis revealed that only sunshine duration significantly increased over the study period, with an inter-annual increase of 3.6 h/year (p < 0.01), whereas inter-annual NDVI trends were negligible; (2) the abrupt change detection showed that a major hydro-climatological change occurred in 2004, when abrupt changes occurred in lake volume, water level, and sunlight duration; and (3) the RDA showed that evaporation and temperature were highly correlated with monthly changes in NDVI. At larger time scales, however, water level and lake volume gradually became more important than evaporation and precipitation in terms of their influence on NDVI. These results suggest that water availability is the most important factor in vegetation restoration. In this paper, we recommend a practical strategy for lake ecosystem restoration that takes into account changes in NDVI.