Human history is shaped by landmark discoveries in science and technology. However, across both time and space the rate of innovation is erratic: Periods of relative inertia alternate with bursts of creative science and rapid cascades of technological innovations. While the origins of the rise and fall in rates of discovery and innovation remain poorly understood, they may reflect adaptive responses to exogenously emerging threats and pressures. Here we examined this possibility by fitting annual rates of scientific discovery and technological innovation to climatic variability and its associated economic pressures and resource scarcity. In time-series data from Europe (1500-1900CE), we indeed found that rates of innovation are higher during prolonged periods of cold (versus warm) surface temperature and during the presence (versus absence) of volcanic dust veils. This negative temperature-innovation link was confirmed in annual time-series for France, Germany, and the United Kingdom (1901-1965CE). Combined, across almost 500 years and over 5,000 documented innovations and discoveries, a 0.5°C increase in temperature associates with a sizable 0.30-0.60 standard deviation decrease in innovation. Results were robust to controlling for fluctuations in population size. Furthermore, and consistent with economic theory and micro-level data on group innovation, path analyses revealed that the relation between harsher climatic conditions between 1500-1900CE and more innovation is mediated by climate-induced economic pressures and resource scarcity.