Cytogenetic analysis has been used for several years in attempts to detect the mutagenic effects of suspect agents. Yet there remain many questions related to statistical analysis. While several procedures have been and are being used, many unfortunately are not always appropriate for the types of data actually obtained. Two general analysis approaches are discussed together with ways to insure that the test assumptions are reasonably met. Because assay sensitivity is a major concern, the sample size issue is discussed with examples for specific situations.