Water eutrophication has become a worldwide environmental problem in recent years. Once a water body is eutrophicated, it will lose its primary functions and subsequently influence sustainable development of society and economy. Therefore, analysis of eutrophication becomes one of the most essential issues at present. With the ability to deal with vague and uncertain information, and express knowledge in a rule form, the rough set theory (RST) has been widely applied in diverse domains. The advantage of RST is that it can compress the rule and remove needless features by reduction inference rule. By this way, the rule gets effectively simplified and inference efficiency gets improved. However, if data amount is relatively big, it could be a process with large calculated amount to search rules by looking up tables. Petri nets (PNs) possesses so powerful parallel reasoning ability that inference result could be obtained rapidly merely by simple matrix manipulation with no need for searching rules by looking up tables. In this work, an integrated RPN model combining RST with PN was used to analyze relations between degrees of water eutrophication level and influence factors in the Pengxi River of Three Gorges Reservoir. It was shown that the RPN model could analyze water eutrophicaion accurately and quickly, and yield decision rules for the decision-makers at water purification plants of the water quality and assist them in making more cost-effective decisions.