The purposes of this paper are to provide an explanation of the way in which suppressor variables operate in multiple regression, and to provide illustrations of their effects with actual social science data. Examples are reviewed in which one predictor accounts for more of the variance in the dependent variable when another predictor is included, and in which the last variable to enter in a stepwise regression equation accounts for more variance than the next-to-last variable. Examples include the following: (1) in predicting the number of doctors in an area from the number of hospital beds, the money spent on hospital facilities was a suppressor variable; (2) in predicting rural or urban preferences for location of the future practice of pharmacy students by its relationship with sex, age was a suppressor variable; (3) in predicting reading scores by a personality inventory, personal integration scores were a suppressor variable for the estheticism score; and (4) in the prediction of the larceny rate in certain census tracts, the robbery rate was a suppressor variable for a socioeconomic indicator. Analysis of variance may be used to clarify suppression. (Author/CTM)