Because citizen reports are the primary means that police learn of crimes, calling the police has been called the most important decision in the criminal justice system. One view of citizen-police cooperation contends that citizens report crimes to the police because they perceive the police to be legitimate. How, then, do shocks to institutional legitimacy shape the demand for police services? Analyzing 25 well-publicized cases of police brutality across 22 US cities using differencein-differences analyses and random permutation tests, I find little evidence that police brutality incidents reduce willingness to call 911 to report crimes overall or in Black neighborhoods, contrary to previous empirical work and some theories of citizen-police cooperation. Analyses of google search trends in impacted areas indicate substantial interest in local brutality events, media reports and public opinion data indicate these events reduced police trust, and many incidents resulted in sustained protests and concrete changes in laws and policies about policing. However, demand for policing services-measured by 911 calls reporting assault, burglary, theft, and gunshots-remained remarkably steady. Robustness tests compare acoustic measures of gunshots with citizen reports of shots fired and examine the racial composition of crime reporters before and after brutality events, providing additional evidence of no effect. In the absence of alternatives, citizens continue to call on police intervention to manage crime despite damaged police legitimacy.