In Réunion Island, herbicides are the main pesticides used in sugarcane cropping systems (Antoir et al., 2016; DAAF La Réunion, 2016). Available herbicides registered for use on sugarcane were recently reduced to 8, and should be further restricted to 4 (ANSES, 2023). However, weeds may cause yield losses ranging from 20% to over 50% (Marnotte & Le Bourgeois, 2018), and are considered as a threat by industry stakeholders (Martin et al., 2012; DAAF La Réunion, 2020). In addition, many other factors can contribute to lower yields, like climatic conditions and some management practice (Christina et al., 2021). To choose and prioritize weed management practices, it is necessary to analyse the crop-weed interaction at stake within the cropping system, even more in tropical context, where there are very few references and studies on this subject. An agronomic diagnosis using crop modelling may be a useful approach to provide a better understanding of the relative importance of weed competition among the other factors limiting yield (Doré et al., 1997; Affholder et al., 2003). The objective of the study presented here is to analyse the role of weed-crop interactions on yield in sugarcane cropping systems in Reunion Island. To do this, small plots of 25 m2 each were identified to minimize intra-plot variability but maximize inter-plot variability within 25 farmers' fields, spread over the five sugarcane production basins of Réunion Island. The main potential local limiting factors were listed according to a literature review and surveys from local stakeholders (growers, technicians and scientists). Based on this, different variables (climate, cropping practices, weed development, crop growth) will be monitored in all farmers' plots over the sugarcane cropping cycle. We will then use this dataset to estimate the respective weight of the different factors in each cropping situation using crop modelling. The comparison between the simulated potential yield and the observed one