This paper aims to show confirmatory evidence concerning the instrument for measuring communication effectiveness of the university teachers proposed by Rego (2000). Previous research show that: (1) the construct comprises four dimensions: empathetic behavior, pedagogical conscientiousness, courtesy, "just reading of the texts"; (2) Cronbach Alphas range between 0.85 and 0.92; (3) the four dimensions explain 85% of the variance of a communication effectiveness score given by 132 students to their actual teachers; (4) they also explain 76% of the variance of the same score given by 69 university teachers to their former teachers. In the present study, 131 university teachers and 87 students described the behaviors of a former teacher, and gave him/her a global score for communication effectiveness. Teachers worked at two Portuguese universities and two polytechnics. Students were from the University of Aveiro (Portugal), attending several graduation and undergraduation courses (e.g., management, economics, industrial engineering). Regression analysis was carried out to test how the communication dimensions explain the communication effectiveness. A cluster analysis was also performed: (1) the teachers were clustered according to their scores on the five communication dimensions; (2) six clusters or configurations were extracted; (3) the clusters were compared in what concerns the six communication dimensions and also the effectiveness score. The findings are the following: (1) the four-factor model proposed by Rego fits the data well (e.g., goodness of fit index is 0.89, normative fit index is 0.89, and comparative and incremental fit indexes are 0.94) ; (2) however, it seems more desirable to divide the first factor in two dimensions (supporting behavior; communication facilitation); (3) indeed, in this five-factor model, goodness of fit index is 0.90, normative fit index is 0.91 and comparative and incremental fit indexes are 0.96; (4) the Cronbach Alphas of these five dimensions range between 0.76 and 0.90; (5) the five dimensions explain 81% of the variance of the communication effectiveness score given by the students; (6) they also explain 77% of the variance of the same score given by the teachers; (7) the configurations receiving the highest (lowest) scores on communication effectiveness are those higher (lower) on the five communication dimensions; (8) according to the teachers and the students, their more effective former teachers were those that promoted the students participation in class, facilitated the communication process (e.g., by using a simple speech), were conscientious, were courteous and did not rely just on the reading of the texts (e.g., transparencies) to give the lessons. In general, the instrument reveals good psychometric properties and highlights some avenues for future research in the domain of quality in higher education. A This research was developed under the Research Project POCTI/CED/40265/2001 - "Citizenship Behaviours of University Teachers", supported by the "Foundation for Science and Technology" with the co-support of the Funds FEDER of the European Union.