Recent palaeobiological studies have emphasized the need for interpretations of the fossil record to consider spatial changes in environmental conditions (e.g. topography, climate). Establishing the role the environment plays in determining the distributions of extinct and existing organisms is complicated by biological evolution. Using available observations to ‘see through’ the randomness of biological evolution to determine contributions from environmental change is not trivial because of the sparsity of the fossil record, lack of precise information about rates of evolution, and because we obviously cannot physically re-run the evolutionary history that resulted in modern biodiversity or the fossil record. To address these issues, we establish scales and scenarios in which spatial environmental change is manifested in records of the number of species in a given area (richness) generated by eco-evolutionary simulation. Evolutionary processes that are likely to be random on the timescales of environmental change are included. Signals of environmental change that are likely to be hidden by the effects of ‘noisy’ evolutionary processes and those likely to emerge are identified. The ‘experiment of life’ is simulated many times, producing statistical insights. Results show that the spatial rate of environmental change is strongly correlated with species richness when the ability of organisms to disperse is high. Interaction between scale, dispersal and environmental structure is shown to determine both statistical and spatial distributions of richness. As a proof-of-concept, we compare predictions to bird species richness. The results emphasize the need to consider the randomness of evolution when interpreting the observations of extinct or present life on Earth.