The relative contribution of modeling and instructions as methods for training counselors in nonverbal empathy was assessed in a 2 X 3 factorial design employing two instruction conditions (Instructions and No Instructions) and three modeling conditions (High Nonverbal Empathy Model, Low Nonverbal Empathy Model, and No Model). Six groups of subjects were presented with the six different combinations of the above conditions and were then asked to conduct a 15-minute interview with a "client." It was predicted that those viewing the High Nonverbal Empathy Model would exhibit higher nonverbal empathy responses in the interview. It was also predicted that instructions would have no effect. The results indicated that these hypotheses were accurate in that there was a significantly higher amount of nonverbal empathy produced subsequent to subjects viewing the High Nonverbal Empathy Model as opposed to the Low Nonverbal Empathy Model or No Model. Instructions had no effect. This lends support to the importance of modeling as a means of teaching nonverbal empathy to counseling students.