This paper analyses the efficient management of nonpoint source pollution (NPS) under limited pollution control budget and incomplete information inherent in NPS pollution. By incorporating information acquisition into a pollution control model, it focuses on the tradeoff between data collection and treatment efforts and derives conditions under which (i) a favourable change in the state of treatment cost at one site may lead to an increase in treatment level at another site, (ii) a higher data collection cost induces more data collection, and (iii) an increase in information productivity leads to an increase in the level of data collection. A numerical simulation of the model illustrates how in managing NPS pollution the value of information acquisition depends on the degree of heterogeneity of polluting sites.