The relevance of ethnographic research to evaluation of educational programs is discussed. The author focuses on evaluation questions that deal with finding out what is happening in a program and if what is happening is desirable. Ethnography suits the purposes of evaluators who answer such questions because they want to know how something works under the worst of conditions. Issues related to the validity and reliability of measurements used in evaluation illustrate this point. The internal validity of a study can be threatened by history, maturation, instrumentation, selection, mortality, and reactive arrangements--all part of the life of the school. To conduct an evaluation by controlling these variables is to study a version of the school setting that can never be replicated. Educational evaluations must be capable of focusing on the program as it really functions. Conventional measures of reliability in evaluations call for inter-subjective agreement. Ethnography, however, focuses on another form of reliability called "confirmability" where there is agreement from a number of information sources that a certain phenomenon has occurred. Finally, the author maintains that the success of an evaluation plan rests on its credibility. Naturalistic inquiry, as structured under ethnography, sustains credibility because of its logic, responsiveness, and involvement of participants. (LC)