BACKGROUND: Randomised controlled trials (RCTs) have been increasingly used to test the effectiveness of mental health and psychosocial support(MHPSS) interventions for populations affected by humanitarian crises. Process evaluations are often integrated within RCTs of psychological interventions to investigate the implementation of the intervention, the impact of context, and possible mechanisms of action. We aimed to explore limitations and strengths of how process evaluations are currently conceptualised and implemented within MHPSS RCTs specifically. METHODS: In April-June 2021 we conducted semi-structured interviews with 24 researchers involved in RCTs of MHPSS interventions in 23 different countries. Participants were selected based on systematic reviews of MHPSS interventions, funders' databases, and personal networks. Data were analysed using codebook thematic analysis. RESULTS: The conduct of process evaluations was characterized by high heterogeneity in perceived function, implementation outcomes assessed, and methods used. While process evaluations were overwhelmingly considered as an important component of an RCT, there were different opinions on their perceived quality. This could be explained by the varying prioritization of effectiveness data over implementation data, confusion around the nature of process evaluations, and challenges in the collection and analysis of process data in humanitarian settings. Various practical recommendations were made by participants to improve future process evaluations in relation to: (i) study design (e.g., embedding process evaluations in study protocol and overall study objectives); (ii) methods (e.g., use of mixed methods); and (iii) increased financial and human resources dedicated to process evaluations. CONCLUSION: The current state of process evaluations in MHPSS RCTs is heterogeneous. The quality of process evaluations should be improved to strengthen implementation science of the growing number of evidence-informed MHPSS interventions.