Abstract: According standard of seawater quality GB3097-1997, plenty of random samples in the concentration scale of different water quality kinds were produced by VB program, then they were used to train and test the artificial neural networks model based on BP arithmetic, and the seawater quality of Pearl River Estuary in 2002 and 2003 were assessed by the BP artificial neural networks model which had been trained. The results showed that the BP artificial neural networks model which had been trained could estimate the seawater quality kinds of unknown seawater samples exactly, and the seawater quality of Pearl River Estuary in 2002 and 2003 were bad, from class II to class IV in most area. On the distribution of the seawater quality in Pearl River Estuary, the western was better than the eastern what was caused by (1) the large amount polluted water discharged from Dongguan and Shenzhen city located at the east seacoast of Pearl River Estuary; (2) the busy sea transportation in east area of Pearl River Estuary.