Objective The attempt of a woman to deliver vaginally after having had a caesarean in a previous pregnancy is increasingly common in current obstetric practice. During a trial of labour after caesarean, gynaecologists consider whether continuing vaginal birth is safe or, alternately, whether a repeat caesarean is advised. There is large variation in the success rates of women with comparable medical risk factors, requiring better insight in how this assessment is made. As a window of opportunity to intervene in this unexplained variation in practice in specific, and in the globally rising caesarean rate in general, our aim was to increase understanding of gynaecologists’ decision-making during trial of labour. Study design We conducted a constructivist grounded theory study, interviewing Dutch gynaecologists. Data collection and analysis were performed concurrently. Initial convenience sampling shifted to theoretical sampling as the study progressed. Data collection continued until theoretical sufficiency was reached. We applied open and axial codes to transcripts of the interviews, and then assembled the axial codes into themes that built up to an emerging theoretical framework. Results Nine gynaecologists were interviewed. Data indicated they continuously weighed the chance of a successful outcome of trial of labour against the likelihood of adversities. Patients’ opinions, aspects of progress of labour and gynaecologists’ personal stances regarding trial of labour played a role in the decision-making process; these factors are influenced by organisational affordances and culture. Variation in the assessment of individuals’ chances of success and variable thresholds for a repeat caesarean added to the complexity of the decision-making. Conclusion This study pieced together patient-, delivery-, physician- and society-related factors that result in vitally important decisions during trial of labour after caesarean; it reveals the complexity as well as the repetitive patterns involved in this process. Exposing these factors offers opportunities to incorporate the decision-making process in targeted educational interventions, with the aim of modifying the underlying assumptions and concepts in order to reduce practice variation.