Wavelet Analysis is a new technique which developing quickly in time-frequency analysis.Because of its characteristics of multiple timescales,multiple layers and multiple resolutions,wavelet analysis has been applied widely in signal analysis,information analysis and earth sciences,etc.In this paper,wavelet technique was used to analyze phytoplankton variation for the first time on the basis of two 30-day in situ observations which carried out in spring and autumn 2002,at the fish-farming area of Dapang'Ao Cove,Daya Bay.Results showed that,the wavelet transform of spring and autumn phytoplankton density time series have characteristics of dominance periods and multiple timescales,which consistent on large scale but differential on small scale.And the analysis results of different wavelet transforms (such as Mexh wavelet transform and Morlet wavelet transform) have some different characteristics.The preliminary application of wavelet analysis technique in ecological study on coastal waters shows that,wavelet analysis technique can be applied to analyze the dynamic variation of phytoplankton in ecosystem on multiple timescales,multiple layers and multiple resolutions,and provids a new method for further studying and forecasting phytoplankton dynamics.