Over the last decade, societal pressure has increased on public research agencies to provide evidence that their research contributes to achieve development goals. To address this challenge, the French Agricultural Research Center for International Development (Cirad), developed an ex-post impact assessment method based on the impact pathway approach, called Impact of Research in the South (ImpresS). The reconstruction of the impact pathway identifies and assesses research outputs, outcomes, and impacts on development over long-time spans, taking into account the contribution of other stakeholders, projects, and contextual factors. By applying mixed methods and participatory approaches, ImpresS involves key actors in assessing the contribution of research to impacts in innovation processes. Such a participatory approach raises, however, questions about the advantages and disadvantages of participation in impact assessment. This article examines whether and how participation affects the results of an evaluation and the methodological challenges this poses. The analysis is based on 13 case studies covering different innovation processes, countries, and time spans. The main results show that participation, combined with triangulation of information including quantitative and qualitative data, strengthens the explanation of the causal relationships among outputs, outcomes, and impacts. It helps reveal a large and diversified list of impacts based on the perception of actors, especially uncovering positive and negative impacts unexpected by researchers. However, participation may render other impacts less visible, especially those related to environmental and political issues. Furthermore, participatory evaluation entangles challenges linked to expectations, divergent perceptions, power dynamics, and social inequality, which must be carefully addressed to provide robust and transparent evaluation results.