The relationship between empathy and sex similarity of counsellors and clients was investigated in this study. Two objectives were advanced to determine: (a) if counsellors of moderate or low empathic ability were more empathic with clients of the opposite sex, and (b) if counsellors of high empathic ability could be equally empathic with clients of both sexes. The subjects were 96 Caucasian University of British Columbia students enrolled in education or counselling psychology, or volunteering at a campus crisis and information centre. They were tested on the Discrimination Empathy Test, a lU item, sexually unbiased revision of Carkhuff's (1969) Index of Discrimination. Their scores were ranked and the subjects designated as of high, medium, or low empathic ability. This designation and the sex of the subjects and the clients served as independent variables. The dependent variable was empathy, defined as the discrimination of client affect. It was measured by the Affective Sensitivity Scale, an empathy measuring instrument consisting of video-taped vignettes from real counselling sessions with clients of both sexes and a written scale on which the subjects recorded their empathic response. Seven null hypotheses were advanced and analysed by a three-way analysis of variance. Only two of the hypotheses were not rejected. Females were found to be more empathic than males, and subjects were more empathic with male than female video-taped clients. For the third main effect, it was found that the three designated levels of empathic ability were not significantly different when re-tested on the Affective Sensitivity Scale. Nonsignificant results were found in interaction analyses of the three independent variables. Speculation was offered to explain these results. The lack of significant differences between the levels of empathic ability was probably responsible for the nonsignificance of three of the four interactions. The fourth, the finding of no interaction between the sex of the subjects and the sex of the video-taped clients could have been produced whether or not similarity was an important factor. The findings of additional t tests between each subject/video-taped client match by sex did not clarify any further the role of similarity. The highest mean score for the matches was for female subjects/ male video-taped clients, followed, in order, by male subjects/male videotaped clients, female subjects/female video-taped clients, and male subjects/ female video-taped clients. The same sex matches were not significantly different. Further research was recommended, especially into the relation of similarity to empathy. / Education, Faculty of / Educational and Counselling Psychology, and Special Education (ECPS), Department of / Graduate