Participatory approaches are widely used in ecosystem services (ES) research. They are particularly advocated for situations characterized by complexity, uncertainties and multiple values. However, behind the intention to do participatory research on ES, there is likely a wide range of practices. In this paper, we undertook a systematic literature review to examine how participatory ES research is implemented in practice. Drawing on 93 reviewed articles, we explore how – and how far – various practices elicit and consider different types of uncertainties related to ES, namely ethical uncertainties (plurality of worldviews, values and interests), epistemic uncertainties (multiple representations) and radical uncertainties (unpredictability). Our review shows a high level of diversity of methods within participatory ES research. Three main types of studies were identified: (1) those centered on socio-cultural valuation of ES, that acknowledge plurality of specific values; (2) those describing more scientific driven processes focusing on assessments of representations of ES dynamics, that partially acknowledge epistemic uncertainties; and (3) those (less numerous) describing more deliberative and collective processes, that navigate all uncertainty types, including plurality of interests, plurality of knowledge systems and radical uncertainties. In total, three main conclusions are drawn from this work. First, plurality of worldviews is seemingly not a strong concern for participatory ES research. This lends credence to concerns that ES framings may encourage a dualistic, anthropocentric and utilitarian framing of nature. Second, although a plurality of specific ES values were generally considered, conflicts of interests and trade-offs between these were much less often considered, which potentially reflects a lack of connection of participatory ES research to real life decision making and a limited ability to navigate power asymmetries and strategic political agendas. Third, while there was often appraisal of non-scientific stakeholders’ representations of ES dynamics, radical uncertainties and differences between scientific and non-scientific representations were rarely addressed. This suggests that participatory ES research remains largely anchored in a Western science’s positivist stance.