The 2019 London Extinction Rebellion was the first attempt by environmental protesters to create prolonged large-scale disruption in a Western capital city. The effects on public opinion were difficult to predict because protests seen as extreme can reduce support, but protests seen as justified can increase support. We studied longitudinal opinion changes in a nationally representative sample (n = 832) before, during, and after the rebellion, in conjunction with experimental analysis of the causal effects of media reports (n = 1441). The rebellion was longitudinally associated with national increases in environmental concern, and activist media increased dissatisfaction with current government action. Reports from an activist, the BBC, and the Daily Mail caused activism intentions and support to move in different directions, contributing to longitudinally increased polarisation in attitudes to activism. The rebellion had minimal effects on belief in whether ordinary people can produce relevant change (based on collective efficacy and support for a Citizens’ Assembly). The rebellion thus succeeded in some aims and did not show evidence of any important backlash.