Anticipating the effects of global change on biodiversity have become a global challenge that requires new methods. Approaches like Species Distribution Models have important limitations which have fuelled the development of Joint Species Distribution Models (JSDM). However, JSDMs rely on community data from structured surveys. Nonetheless, no assessment on the suitability of JSDMs to work with unstructured data from opportunistic databases has been performed. Here we test JSDMs performance when using opportunistic databases. Using artificial data that mimic the limitations of such databases by subsampling complete cooccurrence matrices, we analysed how the completeness of opportunistic databases affects JSDMs in terms of (a) the role of independent variables on species occurrence, (b) residual species cooccurrence (as a proxy of biotic interactions), and (c) species distributions. Moreover, we illustrate how to evaluate completeness at the pixel level of real data with a study case of forest tree species in Europe, and evaluated the role of data completeness in model estimation. Our results with artificial data demonstrate that decreasing the retention percentage increase false negatives and negative cooccurrence probabilities, leading to loss of ecological information. However, JSDMs support different levels of degradation depending on the aspect of the model being considered. Models with 50 % of missing data are valid for estimating species niches and distribution, but interaction matrices would require more complete databases with at least 75% of data retention. Furthermore, in most cases JSDMs predict the original data even better than the data from subsampled matrices, both from testing and training subsets. All those findings were confirmed in the analysis with the real study case. We conclude that opportunistic databases are a valuable data source for JSDMs, but their use require a previous analysis of data completeness for the target taxa in the study area at the spatial resolution of interest.