This study (a post-intervention assessment) was designed to assess the effectiveness of peer-assisted learning (PAL) using student-teachers (STs) with limited training to teach complicated technical skills for interpreting ultrasound images of the shoulder. Students in Years 3 and 4 of medical school were randomly assigned to two groups. In the PAL group (PG), teaching was delivered by a group of nine STs from Years 3 and 4, who undertook a 30-minute general training and 1 week of self-teaching. In the staff-led group (SG), students were taught by a group of three ultrasound-experienced doctors. Exposure took place in two separate lessons (each of 120 minutes) and introduced eight standard sectional planes (EULAR) using a 10-MHz Nemio XG system (Toshiba Medical Systems GmbH). The theoretical and practical learning outcomes were tested using a multiple-choice question (MCQ) test and an objective structured clinical examination (OSCE). Qualitative differences were evaluated using Likert scale-based items. Evaluation of differences between the PG (n = 75) and SG (n = 76) in the theoretical (MCQ score; P = 0.644) and practical (total OSCE score; P = 0.133) outcomes showed no difference between the two groups. However, the STs themselves showed significantly better results overall (P < 0.05). Staff members were rated more highly than STs, especially on items relating to competence (P < 0.05). Complicated technical skills can be adequately taught to students using the PAL system by STs with limited training. Self-teaching learning strategies are successful in contexts of limited teacher training. However, despite positive objective results, STs still face prejudice from students with regard to competency.