Evidence suggests that individual difference measures of personality are limited in their ability to predict cross-situational behavior. This limitation has applied to measures of communication-bound anxiety that have generally been developed as unidimensional instruments. A study explored whether a measurement approach based on individual and situational differences would yield significantly better cross-situational predictions of communication apprehension than would current measurement approaches. In the first phase of the study, an individual difference measure and a situational difference measure were developed and tested with 71 college students. In the second phase, the two instruments were compared with current instruments to determine how well each could predict the same cross-situational behaviors. All instruments were administered to the same group of students. Results showed that the new instruments predited as well as or better than any of the current instruments across a number of situations, with cross-situational predictions of behavior improving from an average of 5.9% for the current measures to 50.4% of the new ones. (Copies of the instruments developed for the study and related materials are appended.) (FL)