<p>Comprehensive assessments of runoff evolution are scientifically essential for effective water resources management under the increasing coupled climate–human effects. Focuses on the Poyang Lake Basin (the largest freshwater lake in China), this study conducted a multi-temporal analysis of runoff evolution at annual, seasonal, monthly, and daily scales during 1953–2022. Base on runoff data from six hydrological and precipitation data from twenty-four meteorological station, methodologies were selected to capture scale-specific characteristics: the Mann–Kendall and Pettitt tests were applied to detect inter-annual trends and abrupt points, the Indicators of Hydrologic Alteration (IHA) were used to examine multi-temporal flow characteristics, and the Slope Change Ratio of Cumulative Quantity (SCRCQ) method quantified the relative contributions of climate change and human activities. Our results demonstrate pronounced scale-dependent behavior in runoff variability. Specifically, no significant trends were detected at annual scale, whereas variability at seasonal and monthly scales increased. Accordingly, three distinct periods were identified: a baseline period (1953–1968) and two altered periods (1969–1992, 1993–2022). Seasonal and monthly runoff evolution during the altered periods varied across the six stations, with strong basin-wide variability in Period 2. At daily scale, hydrological indicators revealed notable changes in flow magnitude and extreme frequency. Precipitation remained the dominant driver, but human impacts intensified since 1969. By integrating analyses across multi-temporal scales, this study provides a multi-temporal research framework that not only facilitates the comprehensive understanding of runoff evolution but also offers valuable insights for adaptive water resource management.</p>