Research into human mobility has exploded in recent years, but work tends to be constrained along 2 key dimensions: spatial context and temporal length, with studies often limited to a few cities or a few months, or both. To understand the durability and generality of many of the patterns we see in mobility research, here we take a wide aperture, using data from the US, UK, Spain, Brazil, Indonesia, India, and Mexico; we also look at some of the longest time series available, constructing panels with as much as 6 years of data. We focus on our propensity to interact with others like us in day-to-day activity—"experienced segregation"—as well as other aspects of urban life. Our results suggest that mobility patterns display variability, drifting over time and differing across space. We examine an important and understudied threat—relative to shocks like COVID—to mobility, which may influence these trends over the next few decades: heat. In various contexts, we show that high temperatures are associated with reductions in activity—and vulnerable groups like older, poorer populations suffer the worst effects. We conclude with the ways that the heat costs urban economies, from harder costs like missed work to softer ones like reduced intergroup interaction.