The Incidence Function Model (IFM) has been put forward as a tool for assessing conservation plans. A key benefit of the IFM is low data requirements: widely available species occurrence data and information about land cover. Citizen science is a promising source of such data; however, to use these data in the IFM there are typically two problems. First, the spatial resolution is too coarse, but existing approaches to downscaling species data tend not to extend to patch level (as required by the IFM). Second, widely available citizen science data typically report species’ presences only. We devise ten different downscaling methods based on theoretical ecological relationships (the species–area relationship and the distance decay of similarity), and test them against each other. The better performing downscaling methods were based on patch area, rather than distance from other occupied patches. These methods allow data at a coarse resolution to be used in the IFM for comparing conservation management and development plans. Further field testing is required to establish the degree to which results of these new methods can be treated as definitive spatially-explicit predictions. To address the issue of false absences, we present a method to estimate the probability that all species have been listed (and thus that a species’ absence from the list represents a true absence), using the species-accumulation curve. This measure of confidence in absence helps both to objectively identify a habitat network for fitting the IFM, and to target areas for further species recording.