an expert system shell called calex has recently been developed for crop management decision support. this shell provides for efficient data entry and management and permits the integration of information from a variety of sources to develop crop management guidelines. the structure of the calex knowledge base language makes it particularly appropriate for complex problems like irrigation scheduling, and the expert system shell was designed with applications as this in mind. the first large-scale implementation of a calex-based crop management decision support system is calex/cotton for irrigated crop management in the san joaquin valley, california the irrigation scheduling module of calex/cotton can use a variety of methods to generate either the estimated date of the next irrigation or, at the user's request, a complete irrigation schedule for the remainder of the season. the methods used include scheduling based on leaf water potential, growing degree-days, the water budget method, last irrigation based on estimated cutout date, and last irrigation based on observed optimal dates. we give here a brief summary of the functioning of the calex expert system shell and then describe in detail the methods of irrigation scheduling used in calex/cotton and how the shell is coupled with appropriate knowledge bases to form an expert system for implementing these methods. (author abstract) ref