Abstract: Upon the 1951-2015 observation runoff data of the five stations in the Southern Jingjiang River section (of the Yangtze River), the present study analyze the section historical runoff evolved characteristics through the methods of MarkKendall Test, Cumulative Anomaly, and Complex Morlet Wavelet Transform, respectively; this study also forecast the future period (2016-2030) runoff changing scenario with the ARIMA model and Time series model. The results show that: (1) Historical annual runoff of this section changed greatly with large discrete degrees, and exhibited significantly seasonal variations, the dry season covered between each November and next April, but the wet season distributed unevenly in other 6 months; (2) This section historical runoff presented a general declining trend over the past 65 years. The largest runoff declining period was 1959-1980 and the trend magnitudes absolute value p was 698.313. The MarkKendall test showed there were no obvious runoff changes in the period of 2003-2015, but the trend magnitudes absolute value p also reached 166.524; (3) The MarkKendall test and the Cumulative Anomaly results showed the sudden change year of this section runoff was 1970 and 1985, respectively; (4) Over the past 65 years, the river section runoff presented three main changing periods, 48-58 a, 20-28 a, and 10-18 a, respectively; the wavelet variance showed the order of three main changing periods was 55 a (first), 24 a (second), and 12 a (third), respectively; (5)This section forecasted future (2016-2030) runoff will be decreased earlier and increased later relative to current conditions, which mean 2019-2026 may be dry period, and 2026-2030 may be wet period, respectively.