This article presents a phenomenological study that was conducted as a series of semi-structured qualitative research interviews with eight student therapists, just two months after their debut in this role. Thematically, the interviews focused on specific examples of good and bad supervisory events. Also included were the importance of peers in the supervision group and the organisational setting of the supervision. The objective was to give detailed descriptions in the form of condensed narratives of each student’s preferences concerning supervision. Furthermore, a cross-sectional analysis was conducted, in order to determine what typically characterized this sample. As expected, we both found major individual differences, but also similarities within and across the sample that confirm previous findings in the literature. With respect to the latter, our findings are very close to those in the existing literature in the field. The beginning therapists prefer supervision in which advice and clear and specific instructions are given on how to do the job, where theoretical considerations are included, and the supervisor supports, affirms and structures the sessions. However, of particular interest are the major individual differences that emerged. These are conspicuous in the presented narratives of each student therapist. Thus, to some therapists, seemingly frustrating supervisory experiences turned out to be very helpful. Role modelling and the parallel process also were found to be of importance for learning. A practical implication of these differences is that supervisors should not merely follow the general directions given in the literature for supervising beginning therapists, but need to adapt at least partly to each supervisee in order to create an optimal learning experience. The limitations of this study are the relatively small sample and the fact that the interviewers were part of the institution in which the supervisory practice takes place, causing at least some students to be restrained in their criticism.