The recent history of crisis intervention training programs for police are summarized, and efforts to evaluate these programs are critiqued. The use of simulation methodologies, such as role playing, to assess intervention skills in a variety of human service fields is reviewed. The design and construction of the CISS are then described in detail. The CISS presents four types of crisis situations, accompanied by a response sheet with open-ended questions and bipolar, Likert-type emotion scales. Results from a pilot study evaluating a 4-day police crisis intervention training workshop using CISS are presented. These results suggest that the CISS was sensitive to changes over the course of the workshop in response to some situations, especially to increases in the reported intensity of certain feelings, and that CISS was able to detect significant relationships between these changes and other variables, including role playing involvement. Suggestions are made for future study into the situation-specific nature of crisis intervention and for evaluation of a modified CISS, in terms of accuracy and effectiveness. Sample situations from the CISS, the response sheet, the critical incident survey form, and a list of 52 references are appended. (Author abstract modified)