This paper examines some of the problems involved in measuring and evaluating peasant farmers' knowledge of their environment and is an extension of the work of Barker. Oguntoyinbo and Richards (1977). That study presented a case for the utilization and inclusion of the perceptions of peasant farmers in the process of monitoring environmental change. Here, the problems of eliciting and evaluating this type of information in methodologically sound ways are discussed, drawing on earlier work in both ethnoscientific Third World studies and conventional behavioural research.