The purpose of this study was to determine whether or not certain attitudes can be modified through laboratory procedures using positively evaluated words within a classical conditioning paradigm. It was hypothesized that evaluations of Vietnamese and Negroes would become more favorable after pictures of each were paired with the presentation of positively evaluated adjectives. Evaluations were made through the use of a semantic differential-type attitude scale. Analysis of covariance demonstrated a significant increase in favorable evaluations toward the Vietnamese only. The reason the evaluation toward the Negroes did not demonstrate similar change may have been due to the fact that Caucasian subjects' attitudes toward Negroes are too psychologically significant an issue to be modified by a single session of conditioning trials. (Author/DP)