The extent of the River Wisla (Vistula) water plume in the Gulf of Gdansk was investigated using SeaWiFS (Sea-viewing Wide Field-of-view Sensor) data as the basis for PCA (Principal Components Analysis), indexation, composite and classic methods of classification (i.e. maximum likelihood classification and fuzzy sets). The percentage transformation of this data is suggested in order to obtain a high relative diversity of remotely sensed signals emerging from optically different waters. Comparison of the results obtained using these different classification methods showed that PCA was especially useful in this respect.