The alarming rate of biodiversity loss worldwide has increased the need for high-quality occurrence data to support conservation assessments for the IUCN Red List of Threatened Species. Plants, in particular, are underrepresented on the Red List. Narrow-ranged endemic plant species from tropical forests, which are highly vulnerable to climate change and anthropogenic activities, require our immediate attention but pose specific challenges as they frequently suffer from a lack of data. As they are often only known from a handful of preserved specimens in herbaria, maximum efforts need to be made on the completeness and correctness of the data to increase their usefulness for research and conservation. Two spatial metrics form the basis for most plant risk assessments: the extent of occurrence (EOO) and the area of occupancy (AOO). Together with information about habitat, elevation and local threats, among others, a species can be assigned to a particular risk category when these two quantitative measurements fall within the relevant thresholds. When the geographic coordinates of some specimens are missing, one should consider georeferencing them before calculating these range statistics. Especially for species with few occurrence records, the inclusion of newly georeferenced specimen data may easily cause the calculated EOO or the number of threat-defined locations to exceed one of the IUCN thresholds, thus altering the conservation status. Improving data quality is equally important to obtain the most complete and accurate occurrence data for a certain species. As the EOO is particularly sensitive to geographic outliers, taxonomic and spatial errors could lead to a miscalculation of the extinction risk, especially for threatened taxa. We illustrate some of these issues and the level of data cleaning needed when using natural history collections for Red Listing poorly-sampled endemic and range-restricted species. We do this using a specimen-based occurrence dataset compiled in the framework of the Conservation of Endemic Central African Trees (ECAT) project, which produced Red List assessments for 347 tree taxa endemic or sub-endemic to the Central African region comprising the Democratic Republic of the Congo, Rwanda and Burundi.