Abstract: This paper quantifies the chaos characters of annual flow time series for the Yangtze River, deals with Lyapunov exponents and their practice meaning, calcu ates the long-term correlation exponent, i. e. tractional difference exponent, and discusses the possible relation between the exponent and the value of the correlation fractal dimension. Theses nonlinear information can reveal the regularity for the annual flow time series and advance the prediction level.