Matrices or look-up tables are increasingly popular flexible tools for ecosystem services mapping and assessment. The matrix approach links ecosystem types or land cover types to ecosystem services by providing a score for ecosystem service (ES) capacity, supply, use, demand or other concepts. Using expert elicitation enables quick and integrative ES scoring that can meet general demand for validated ES mapping and assessment at different scales. Nevertheless, guidance is needed on how to collect and integrate expert knowledge to address some of the biases and limits of the expert elicitation method. This paper aims to propose a set of guidelines to produce ES matrices based on expert knowledge. It builds on existing literature and experience acquired through the production of several ES matrices in several ES assessments carried out in France. We propose a 7-steps methodology for the expert-based matrix approach that aims to promote cogency in the method and coherency in the matrices produced. The aim here is to use collective knowledge to produce semi-quantitative estimates of ES quantities and not to analyse individual or societal preferences or importance of ES. The definition of the objectives and the preparation phase is particularly important in order to define the components of capacity to demand ES chain to be addressed. The objectives and the ES components addressed will influence the composition of the expert panel. We recommend an individual ‡ ‡ © Campagne C, Roche P. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.filling of an empty matrix in order to strengthen the statistical analysis of the scores' variability and the analysis of congruency between experts. Expert scoring should follow a process of discussion, information-sharing and collective appropriation of a list of ecosystem types and ES to be assessed. We suggest that the ES matrix should not only focus on ES central scores but also address the variabilities and uncertainties as part of the ES assessment. The analysis of these sources of variability allows the documentation of variations in the ES quantity but also an exploration into the lack of consensus or knowledge gaps that needs to be addressed.