Temperature has strong impacts on all biological and ecological processes, and thermal performance curves (TPCs) have been employed recurrently to assess them. TPCs almost always take a particular asymmetric shape across the biological hierarchy, with many different competing mechanisms and models doing a similarly good job of trying to explain the TPC phenomenon. Here, we reveal that the ubiquitous exponential scaling of biological processes with temperature creates a mechanistic tendency for TPC data and models to collapse onto a single curve (which we call the Universal TPC, UTPC), explaining mathematically why biological systems respond to temperature in such a consistent way. We illustrate that many seemingly different TPCs actually approximate rescaled versions of the same curve, even when thermal performance estimates vary widely across organisms, systems, and contexts. We demonstrate remarkable UTPC collapse across the tree of life, with diverse datasets spanning microbes to vertebrates, and individual physiology to population growth. UTPC phenomena also provide a strong theoretical basis for predicting performance of warm-adapted organisms will be more sensitive to- and less tolerant of- temperature fluctuations; an important consideration in the context of climate change.