School-leavers (10) were allocated either to an interview-training group (E) or a discussion control group (C) for 3 sessions. For E group a combination of modeling, coaching, roleplay, feedback and discussion was used to train both verbal and non-verbal interview skills. An attempt was made to program generalization of treatment effect into training and subjective evaluations of interview skill were made by an experienced personnel officer to assess the social importance of the skills trained. After C group had completed 3 sessions to control for the non-specific effects of being in therapy, they received the same training as E group. Subjects were assessed using videotaped roleplayed interviews at the beginning and end of each training phase. After training, E group showed significant improvements of both global and specific ratings of interview skill compared to C group; there was evidence of generalization and maintenance of treatment effect and the social importance of the skills trained was substantiated by the subjective evaluation data. C group then went on to replicate the changes in E group.