Soil is fundamental for the functioning of terrestrial ecosystems, but our knowledge about soil organisms and the habitat they provide (shortly: Soil biodiversity) is poorly developed. For instance, the European Atlas of Soil Biodiversity and the Global Soil Biodiversity Atlas contain maps with rather coarse information on soil biodiversity. This paper presents a methodology to map soil biodiversity with limited data and models. Two issues were addressed. First, the lack of consensus to quantify the soil biodiversity function and second, the limited data to represent large areas. For the later issue, we applied a digital soil mapping (DSM) approach at the scale of the Netherlands and Europe. Data of five groups of soil organisms (earthworms, enchytraeids, micro-arthropods, nematodes, and micro-organisms) in the Netherlands were linked to soil habitat predictors (chemical soil attributes) in a regression analysis. High-resolution maps with soil characteristics were then used together with a model for the soil biodiversity function with equal weights for each group of organisms. To predict soil biodiversity at the scale of Europe, data for soil biological (earthworms and bacteria) and chemical (pH, soil organic matter, and nutrient content) attributes were used in a soil biodiversity model. Differential weights were assigned to the soil attributes after consulting a group of scientists. The issue of reducing uncertainty in soil biodiversity modelling and mapping by the use of data from biological soil attributes is discussed. Considering the importance of soil biodiversity to support the delivery of ecosystem services, the ability to create maps illustrating an aggregate measure of soil biodiversity is a key to future environmental policymaking, optimizing land use, and land management decision support taking into account the loss and gains on soil biodiversity.