Urban micro-farms are rapidly developping driven by NGOs and supported by citizen and local politics. This new object of research was defined in two different contexts: urban and peri-urban/rural environments (Daniel 2019; Morel et al. 2017). In urban environments micro-farm are small scale farms (less than 1.5EFT per ha), involving often volunteers and generating a diversity of activities. The different functions expected from them (food provisioning, teaching support, social and cultural activity, water retention, valorization of wastes and landscape...) are directly linked to a diversity of ecosystem services that they can deliver. Nevertheless, only a few studies focus on micro-farms and do not allow a sufficient understanding of the ecosystem services that they can deliver. In parallel, the involvement of farmers in the evaluation of ecosystem services seems to be a great opportunity, if not a necessary condition. Based on this observation, the research project SEMOIRS (2018 – 2020), funded by ADEME, aims to evaluate ecosystem services delivered by urban micro-farms. The aim of this article is to present and discuss the methodology of the project. Seven micro-farms were studied during two years: 4 were located at the soil level and 3 at rooftop level [see table 1]. Five ecosystem services were studied: food production (quantity and quality), biodiversity supporting (flora, fauna, and microorganisms), regulation (water catchment and quality, local climate regulation, use of urban waste and carbon storage) as well as social and cultural services. A set of indicators was set up [see table 3] based on three types of data : (i) participative measurements achieved by the farmers themselves, (ii) field measurements (including a questionnaire for micro-farm users) performed by researchers and (iii) external data, based on laboratory measurements, data bases, literature etc. The study has shown that all of the seven soils of micro-farms are dee- ply disrupted, from removed topsoil to isolatic Technosol (IUSS Working Group WRB, 2014). The methodology used raised key questions regarding ecosystem services evaluation. (i) First, there is a trade-off to be dealt within the participative approach, between the targeted precision of data needed on the one hand and the time spent by stakeholders for the measurement on the other hand. Despite these limits, the participative approach allows for a better understanding of the ecosystem services drivers by the farmers as well as collecting data not accessible otherwise. (ii) Second, the choice of references to which we compare the ecosystem services delivered by urban micro-farms is crucial. In the project, three reference were considered: professional vegetables farms, other urban agriculture forms and other type of urban green spaces. We identified a serious lack of reference data on the considered ecosystem services that limits the comparison and the discussion. (iii) Finally, our work raised the question of the scale at which to study different ecosystem services, as the different ecosystem services operate at different spatial scales. Four scale were used here: soil, farm, neighborhood and the territory (here, the city; see figure 3). Encompassing different scales raises however the questions of a common study scale and that of the methodology needed to change from one scale to another. However, despite those limits and perspectives the proposed methodology, relying partly on a participatory approach, appears promising to evaluate the ecosystem services provided by urban micro-farms