Purpose We revisit the Rodney King incident and verdict to study their effect on confidence in local police using a design-based causal inference approach. Methods We apply rigorous state-of-the-art quasi-experimental methods to analyze survey data from Southern California and Los Angeles in 1991 and 1992 overlapping with the two focal events. Results While we find a substantial decrease in confidence in the local police both after the incident as well as the verdict, contrary to previous research using non-quasi-experimental designs, our results demonstrate that the loss of confidence caused by the incident varied only modestly by ethnicity and not at all by political orientation. The negative effect of the verdict only varied to a limited extent by political orientation but not ethnicity. Additionally, although there is robust evidence that the incident in 1991 did indeed have a causal negative effect, this evidence is substantially weaker for the effect of the verdict. Given the pre-existing negative time trend prior to the acquittal in April 1992, it is doubtful that the verdict itself played a causal role in mobilizing public opinion. Conclusions Our results shed new light on these two pivotal events and their consequences, which are discussed alongside contemporary research on police-citizen relations.