As more priority is given to improving performance on existing canal irrigation systems, it becomes more important to understand and improve the process of identifying precisely what to do. Many types and combinations of actions are possible, but how to identify which are best has been neglected. Each irrigation system is unique and requires a specially tailored program. Detailed appraisals of the large numbers of existing projects impose impossible demands for high-level specialists and lead to long delays in the delivery of information. Cost-effective approaches and methods are needed which make manageable demands on staff and lead to implementable actions with early benefit. This paper assists in the search for such methods. It draws on experience with rapid rural appraisal (RRA) in other fields1 and on experience with canal irrigation mainly in Asia. Appraisal is used in the general sense to mean investigation and analysis. Primary attention is given to practical investigation, i.e., finding out about existing canal irrigation systems. Diagnostic analytical approaches are treated in detail in a separate paper (Chambers and Carruthers 1985).