BACKGROUND: An international workshop on population health intervention research (PHIR) was organized to foster exchanges between experts from different disciplines and different fields. AIMS: This paper aims to summarize the discussions around one of the issues addressed: the place or role of pilot studies in PHIR. Pilot studies are well-established in biomedical research, but the situation is more ambiguous for PHIR, in which a pilot study could refer to different purposes. METHODS: The workshop included formal presentations of participants and moderated discussions. An oral synthesis was carried out by a rapporteur to validate by expert consensus the key points of the discussion and the recommendations. All discussions have been recorded and fully transcribed. DISCUSSION: PHIR generally addresses complex interventions. Thus, numerous tasks may be required to inform the intervention and test different aspects of its design and implementation. While in clinical research the pilot study mainly concerns the preparation of the trial, in PHIR the pilot study focuses on the preparation of both the intervention and the trial. In particular, pilot studies in PHIR could be used for viability evaluation and theory development. RECOMMENDATIONS FROM THE WORKSHOP PARTICIPANTS: The following recommendations were generated by consensus from the workshop discussions: i) terms need to be clarified for PHIR; ii) reporting and publication should be standardized and transparency should be promoted; iii) the objectives and research questions should drive the methods used and be clearly stated; iv) a pilot study is generally needed for complex intervention evaluation and for research-designed programs; and v) for field-designed programs, it is important to integrate evaluability assessments as pilot studies . CONCLUSION: Pilot studies play an important role in intervention development and evaluation. In particular, they contribute to a better understanding of the mechanisms of intervention and the conditions of its applicability and transferability. Pilot studies could therefore facilitate evidence-based decisions about design and conduct of main studies aimed to generate evidence to inform public health policy.< Réduire