BackgroundTo assess the impact of 10 years of simulation-based shoulder dystocia training on clinical outcomes, staff confidence, management, and to scrutinize the characteristics of the pedagogical practice of the simulation training.MethodsIn 2008, a simulation-based team-training program (PROBE) was introduced at a medium sized delivery unit in Linköping, Sweden. Data concerning maternal characteristics, management, and obstetric outcomes was compared between three groups; prePROBE (before PROBE was introduced, 2004–2007), early postPROBE (2008–2011) and late postPROBE (2012–2015). Staff responded to an electronic questionnaire, which included questions about self-confidence and perceived sense of security in acute obstetrical situations. Empirical data from the pedagogical practice was gathered through observational field notes of video-recordings of maternity care teams participating in simulation exercises and was further analyzed using collaborative video analysis.ResultsThe number of diagnosed shoulder dystocia increased from 0.9/1000 prePROBE to 1.8 and 2.5/1000 postPROBE. There were no differences in maternal characteristics between the groups. The rate of brachial plexus injuries in deliveries complicated with shoulder dystocia was 73% prePROBE compared to 17% in the late postPROBE group (p > 0.05). The dominant maneuver to solve the shoulder dystocia changed from posterior arm extraction to internal rotation of the anterior shoulder between the pre and postPROBE groups. The staff questionnaire showed how the majority of the staff (48–62%) felt more confident when handling a shoulder dystocia after PROBE training. A model of facilitating relational reflection adopted seems to provide ways of keeping the collaboration and learning in the interprofessional team clearly focused.ConclusionsTo introduce and sustain a shoulder dystocia training program for delivery staff improved clinical outcome. The impaired management and outcome of this rare, emergent and unexpectedly event might be explained by the learning effect in the debriefing model, clearly focused on the team and related to daily clinical practice.Electronic supplementary materialThe online version of this article (10.1186/s12884-018-2001-0) contains supplementary material, which is available to authorized users.