Configural Frequency Analysis (CFA) is a method for the analysis of cross‐classifications of categorical variables. Using CFA, one examines individual cells or groups of cells of cross‐classifications with the goal of determining whether they contain more or fewer cases than expected based on some chance model. The specification of the chance model determines the interpretation of results. Depending on chance model, cells with more or fewer cases than expected can be interpreted as indicative of local associations, but also of predictor‐criterion relationships; they allow one to distinguish between groups, or they reflect typical and atypical patterns of change over time.