The paper assumes that the reader is familiar with analysis of variance and with regression analysis. Analysis of covariance is useful when the researcher wishes to examine the relationships between at least two quantitative variables and at least one additional qualitative variable. Analysis of covariance is a special case of the linear model. The volume's first section discusses four applications of covariance in social science research. The next section describes the technique's main uses: to increase the precision of randomized experiments, to remove bias due to nonrandom assignment of experimental test units, to remove bias in observational studies, and to perform regression analysis with multiple classifications. Statistical procedures for analysis of covariance for a one-way layout are next presented. Analysis of covariance for more complex layouts, included randomized complete block design and a factorial design are also presented. Tables, figures, 19 references, and an appendix discussing the model's assumptions and sensitivity are included. For other volumes in the series, see NCJ 72008, 72010-12 and 72014.