<p>We examine the causal effect of temperature on social conflict in China, with a specific focus on labor unrest including strikes and worker protests—which constitutes a frequently overlooked form of social conflict. Using city-daily weather data matched with social conflicts (2014–2020), we employ a high-dimensional fixed effects to account for time-invariant city characteristics and common temporal shocks, thereby isolating plausibly exogenous temperature variation. We find that a one-standard-deviation increase in temperature raises the probability of social conflicts by 0.21 percentage points, representing a 38% increase relative to the sample mean conflict rate of 6%, with effects intensifying during politically sensitive periods and in regions that are more economically vulnerable or have historically cooler climates. Using temperature bins to estimate nonlinear effects, we document that conflict incidence rises sharply at extreme temperatures relative to the reference range (6–10 °C). These findings remain robust across multiple specification checks. Moreover, we explore the role of thermal adaptation, demonstrating that unexpected temperature increases beyond adaptation levels can trigger social conflicts, and these effects are only slightly mitigated by improvements in physiological adaptation. Finally, we provide suggestive evidence that insufficient sleep and heightened negative emotions during hot days serve as potential mechanisms.</p>