A numerical classification of 21 monitoring stations of the lagoonal - estuarine system comprised by the Ciénaga Grande de Santa Marta and Pajarales complex was carried out according to their similarity in water salinity behavior. Biweekly data from January 1987 to Octuber 1991 was used. Due to the autocorrelation structure of the data, it was necessary to use first time series ARIMA models followed by normal cluster analysis of the AR coefficients of infinite representation calculated for each station. The results were coherent with hypotheses concerning the distribution of the variable in the system.