The effects of the meta-model questioning strategy from Neuro-Linguistic Programming were compared to the effects of an empathic responding strategy in the initial counseling interview. Seventy-two clients were obtained from a pool of undergraduate psychology student volunteers. Two graduate students majoring in counselor education served as the counselors and were trained to deliver the two counseling strategies. A four group design was used for the investigation. Clients were assigned at random to one of the two counselors and to one of the two counseling strategies. Each client was treated in one thirty-minute interview;Five dependent variables were identified to assess the effects of the treatments. The first, concreteness in client statements, was determined by trained raters of concreteness using a five-minute transcribed segment of each interview. The second dependent variable was the change in client anxiety during the counseling interview. Client self-reported anxiety change was assessed immediately following the interview. The final three dependent variables were counselor attractiveness, expertness, and trustworthiness, as perceived by the clients. The Counselor Rating Form was completed by each client immediately following the counseling interview to assess these final three variables;No difference could be shown between the two counseling strategies on any of the five dependent variables. Also, no differences could be shown between the two counselors. An interaction of treatment strategy and counselor was found for self-reported change in client anxiety. The counselors, when using either of the two strategies, were generally perceived to be highly attractive, expert, and trustworthy. On the average, client self-reported anxiety was lowered somewhat for both treatments. Client concreteness scores averaged just below the midpoint on a five-point concreteness scale for both treatments. Possible conclusions and recommendations are discussed.