Aim
Energy availability is often assumed to be a primary driver of variation in biodiversity across large‐scale environmental gradients such as elevation, but support for the role of energetic constraints is far from universal. Here, we assume that these variations in energy–richness relationships might arise from the degree of match between the scale of the response of organisms to energy (response scale) and the scale of the observation made (observation scale).

Location
41 mountain ecosystems worldwide, with detailed case studies on Dongling Mountain, China.

Time Period
Contemporary.

Major Taxa Studied
Trees, shrubs, herbs, ferns, birds, ants, bees, Lepidoptera, Coleoptera, Orthoptera, soil fauna, litter‐layer fauna, bacteria, fungi.

Methods
First, we quantify the response scale on the basis of energy requirements, and then, we develop a general scaling model that explicitly describes how the scale matching between response and observation scale influences the predictive power of energy in determining elevational richness patterns; this model holds across taxonomic groups, trophic guilds, and ecosystems across mountains globally.

Results
Global analyses strongly supported the model's prediction that scale matching determines the detectability of energy‐richness relationships. On Dongling Mountain, energy availability best explained richness patterns at the predicted response scales for trees, herbs, and litter fauna.

Main Conclusions
Our model advances the understanding of montane biodiversity distributions by demonstrating the critical role of scale matching in energy–richness relationships. Our findings strongly support the existing mechanistic foundation for predicting biodiversity changes across spatial scales in montane ecosystems. The scale‐matching principle we identified is likely to be applicable to a wide range of taxa and other ecosystems where energy availability varies across scales.