Abstract: Reverberation time series recorded in laboratory cistern, lake field and sea field have been examined by nonlinear dynamical methods, in order to identify chaos characteristic from its random-like waveform and to modeling them by low dimensions nonlinear dynamics. The data we analyzed are related to many geographic locations, bottom types, underwater conditions and transmitted signals. The results indicate that the reverberation can be modeled as 4-dimensions dynamical system with a positive maximum Lyapunov exponent (0 < λ1 < 0.3), so it may reveal a new approach based on chaos to process reverberation time series.